TCP/SRL/6712
Field Document No. 4
TECHNICAL COOPERATION PROGRAMME
BASED ON THE WORK OF
G. J. Meaden
FAO Consultant in Application of GIS and Remote Sensing
FAO - FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS
BANGKOK, 1998
The designations employed and the presentation of the material in this document do not imply the expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United Nations concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.
The Food and Agriculture Organization is greatly indebted to the organizations and individuals who assisted in the implementation of the project by providing information, advice and facilities.
This electronic document has been scanned using optical character recognition (OCR) software and careful manual recorrection. Even if the quality of digitalisation is high, the FAO declines all responsibility for any discrepancies that may exist between the present document and its original printed version.
1. Introduction and Terms of Reference
2. Equipment and Staffing Situation in the Planning and Information Unit
3. Progress Made in Data Collection and Base Map Construction
4. Work Required to Complete GIS Output
5. Acquiring Information to Accelerate the Development of Inland Aquaculture and Fisheries
6. Conclusions and Recommendations
ANNEX 1 Itinerary and Persons Met
ANNEX 2 Resources Used and Methods Applied in Construction of Base Maps for Pilot Area
ANNEX 3 Scoring Procedures to be Followed for "Shrimp Farming"
ANNEX 4 Stages in Combining Map Layers to Produce a Shrimp Aqauculture Zoning Map
ANNEX 5 Work to be Carried Out by PIU in the Immediate Future
ANNEX 6 Plan of Work for the PIU over the Three Months
ANNEX 7 Proposed Amended Terms of Reference to Cover an Extension of the National GIS Consultant's Work on TCP/SRL/6712
ANNEX 8 Bibliography
This report covers the second mission to Sri Lanka by this consultant - a report "TCP/SRL/6712 - Field Document No.2" details progress made during the first visit. Both missions have been concerned with the establishment of a Planning Information Unit (PIU) within the National Aquatic Resources Agency (NARA), within which a Geographic Information System is to be implemented, which itself will allow for the zoning of aquaculture and some planning for inland fisheries rehabilitation. In the original Project Document (1996), it was envisaged that this consultant would make two visits for this TCP, one of about six weeks and the other of seven weeks. Because, by the end of the first mission, progress on the GIS implementation was found to be extremely limited, then it was agreed that the proposed second visit should be split. This report therefore covers a new "second mission" of three weeks duration. Full details of this consultant's itinerary are given as Annex 1.
Working within the original Terms of Reference (TOR's), a set of tasks was proposed to cover the anticipated work which now needs pursuing if the GIS implementation part of this TCP is to be successful. These tasks are:
1. Ensure that all outstanding GIS related equipment for the PIU is in place, and that the staff have suitable experience in any necessary operating procedures (hardware and software).2. Complete the requisite data gathering to allow for aquaculture zoning in the pilot coastal area, and to ensure that PIU staff can convert this data into formats which produce the required GIS output.
3. Obtain some GIS output for the pilot area, i.e. in order to demonstrate the PIU's efficacy to the National Project Co-ordinating Committee.
4. Review the likely success of the pilot area GIS, and then initiate data collection (and digital conversion) for a national coastal aquaculture zoning report.
5. Review the data requirements for an "Inland Aquaculture and Fisheries GIS", and initiate data collection procedures.
Relevant to task 5 above, it was also intended that contact be made with an Australian led group which, working under the auspices of the Australian Centre for International Agricultural Research (ACIAR) and in conjunction with the University of Kelaniya, was to instigate a programme to help revitalise inland fisheries through the use of GIS. This contact was to ensure that some form of co-operative (or complementary) working arrangements could be established.
Under the original plans for this project, it was envisaged that a range of GIS equipment should be purchased through the FAO, and that it should be in place within NARA at the commencement of this consultant's first visit. NARA were to be responsible for providing suitable working accommodation for a unit that would house the equipment, i.e. the Planning and Information Unit (PIU).
Upon arrival at NARA it was clear that very little progress on equipment supplies had been made during the consultant's 7 months absence. Of the 11 items listed in the consultant's first mission report as "awaiting delivery", only three had actually arrived and had been installed. The situation regarding the other items was:
1. Lap Top computer - with MFARD.2. Modem. This should be fitted as a card to the above lap top computer, and should have been supplied with same.
3. A1 Digitiser. The supplier (EMSO Ltd) informed that the digitiser has not yet arrived in the country.
4. Geographical Positioning System (GPS) not yet delivered to the suppliers (EMSO Ltd).
5. 3-way Data Switch. Not yet purchased, it appears that this item was never ordered. This is a fairly inexpensive item.
6. UPS (Universal Power Supply). Not yet ordered.
7. Tape Back-up Drive. This was delivered without the necessary internal card. The PIU are awaiting delivery of this card.
8. Air Conditioning Unit. A unit of the wrong capacity was delivered, but a replacement arrived, though it has yet to be fitted.
Whilst in Colombo, strenuous attempts have been made to resolve this equipment situation by both this author and by the FAO representation. Nevertheless, upon my departure, little further progress had been made on any of the above. On the last day of this mission, the PIU were able to retrieve the lap top computer from MFARD. The lack of access to a digitiser has proven (and is proving) to be a major barrier to progress on this TCP. All digitising to date has been carried out using the facilities of the Mahaweli Project, but this access can no longer be taken advantage of without some compensation to them. There have also been problems with obtaining access to use of the vehicle assigned to this TCP. Steps have now been taken to assure that the vehicle is readily available when required for pre-programmed PIU use, i.e. for trips which have been indicated to the National Project Co-ordinator on monthly work plans.
In order to find out how much the designated PIU staff had learned about GIS hardware, software and requisite processes, and how they saw their role within the PIU as working out, each staff member was briefly interviewed. The results are summarised as follows:
a) Ajith Gunaratne: Has gained a comprehensive knowledge of both of the main GIS software packages and feels confident to use them for a range of tasks. Attended the University of Colombo GIS training course, plus an RS course in Thailand. Completely familiar with all hardware items, and very good on all aspects of computing. Feels that he plays very much an organisational and co-ordinating role. Is able to devote about 4 days per week to PIU/TCP work.b) Ananda Athukorala: Has a comprehensive knowledge of the IDRISI software, and is reasonable familiar with computing demands. Attended the GIS training course at the University of Colombo. Mainly involved in data collection tasks. Has been working an average of 2 days per week on PIU/TCP work.
c) Dr Sepalika Jayamanne: Has some knowledge of the two GIS software packages, but feels that she needs more training, especially on IDRISI. Attended the GIS training course at the University of Colombo. Is fairly competent on the computer. Mainly involved in the collection of data within NARA. Has been working about 1 day per week on PIU/TCP activities.
d) Dr Siriwardena. His other duties prevent him from taking an active role in the TCP work, but he has provided some advisory inputs.
The "Report of the Backstopping Visit for the GIS Component", issued by Dr Kapetsky (April, 1997), strongly recommended that the work input to PIU activities should be based upon full time working by four appointees to a PIU. Over the period since this author's last visit (7 months), it is clear that working productivity inputs have only been at about 35% of the optimum. The fact that a reasonable amount of mapping output and database construction has been achieved (see below), is mainly due to the effort inputs of the National Consultant in GIS (Mr Adikari).
During the mission, this problem of a deficiency of work input was discussed with the FAO representative, with the National Project Co-ordinator and at a meeting of the Project Co-ordinating Committee. NARA have been experiencing some severe difficulties with regard to staffing. This has meant that it has proved impossible for them to release workers to the PIU on a full time basis. I have been assured that plans have now been put into action which will enable an incremental increase in the amount of time devoted to the TCP work. Dr Jayamanne should slowly be able to do more PIU work so that in a few months time this could reach four days per week, and there is a probability that an extra day's work per week could be obtained from Mr Athukorala. If these two targets can be achieved, then actual optimum productivity would be 60%.
The working environment within the PIU facilities is very good. The rooms are adequately furnished, except that a drawer for storing paper maps is required. A quotation is to be obtained for the supply of this. A new air conditioning unit is awaiting installation. The situation regarding connections to the Internet are far from favourable, but this reflects the lack of telephone connection lines to NARA. A small collection of GIS related literature has been started. The staff of the PIU appear to have a very good working relationship.
In this consultant's first mission report, insufficient attention was given to the fact that the "zoning for aquaculture" component of the TCP is being organised as a three stage process. This means that, in order that the methodology being followed can be tested, then zoning is first being carried out in a pilot coastal area. This area includes all the coastline between approximately Chilaw in the south, and the northern section of the Puttalam Lagoon, i.e. on the central west coast of Sri Lanka. The pilot survey is only concerned with zoning for shrimp farming potential. Once the GIS based zoning methods have been tested and adjusted on shrimp farming, then it is the intention to move to the second stage, which is to zone for all other types of coastal aquaculture in the pilot area. The final stage will be to enact refined zoning procedures so as to reveal comparative ratings for aquaculture potential over the whole country. (More facts on this three stage process were given in Annex 4 of the first mission report).
A project which is designed to "zone for coastal aquaculture" must derive a definition of "the coastal zone", i.e. as it applies to aquaculture. Since different forms of aquaculture will be taking place in a variety of environments, and within certain clearly defined parameters, then it is impossible to state a definitive, collective definition of the coastal zone. However, for the purposes of the study, at both the national scale and for the pilot area, we use a definition which encompasses the broadest possible zone in which any form of marine aquaculture might be contemplated. For the pilot study area, the PIU have been drawing up base maps which cover a zone with a width of approximately 10 kms. This may seem to be a wide zone, but allowance must be made for the numerous lagoons which occur around the Sri Lanka coastline, lagoons in which a variety of aquaculture might be pursued. Data to produce all the requisite base maps is to be collected covering the land area in this coastal zone and, for some marine variables, in waters immediately adjacent to this zone.
During the seven months since the first mission, the PIU have collected a significant amount of data. This comprises of published materials, original and photocopied maps, digital datasets, etc. The quantity is too large to itemise here. Most of the effort to date has been directed towards collecting data for the pilot area. However, at the national scale, the PIU has digital vector datasets which allow for the mapping of the following:
a) The Sri Lanka coastline.
b) National Park boundaries.
c) Main roads.
d) Location of tanks.
e) Geological formations (at a very broad classification level).
f) Location of mangroves.
g) Sea bottom sediments in coastal waters.
h) Coastal bathymetry.
i) District boundaries.
For the pilot study area, the data volumes acquired are quite significant, if sometimes fragmented. Upon the consultant's arrival at NARA, the PIU produced A4 size colour print outs of all the digital maps which had been completed, i.e. which were generated from data or datasets collected during the last 7 months. These pilot area maps consisted of:
a) Conflicts in water resources use (In need of minor revisions).
b) Rivers (In need of minor revision)
c) Coast line for the pilot area
d) Bathymetry (In need of major additions)
e) Annual ground water recharge rates
f) Location of hatcheries
g) Urban areas
h) Location of shrimp farms (In need of minor revisions)
i) Power lines and transformers
j) Major and minor roads
k) Areas liable to flooding (In need of minor revisions)
l) Protected areas
m) Mangroves (To be merged with Protected areas)
n) Soils (Will need major revisions)
o) Salinity (Will need major revisions)
All maps were held as digital files within both the Arc/View and IDRISI GIS's. This consultant spent some considerable time reviewing the maps, i.e. checking their validity against data sources, and in some cases checking the veracity of the data sources. For most of the maps little revision was found to be necessary. For some maps (see above) this author suggested minor revisions, and for others (see above) major revisions will be needed. In order that the zoning procedures could be completed in the pilot area, there were a number of maps that had yet to be completed or which had not yet been attempted. These were as follows:
a) Low lying coastal areas.
b) Water quality.
c) Soil pH.
d) Population density
Details on the construction of all the base maps created, as well as those still to be completed, are given as Annex 2. It is important to note here that there are limitations with the quality of some of the data used. For each of the maps described in Annex 2, an attempt has therefore been made to give an indication of the data quality, and where necessary indication of how better data might be obtained in the future.
This consultant, together with this project's International Aquaculture Consultant, spent a considerable amount of time during this mission working on the scoring methodology to plan for shrimp aquaculture, and this proved to be a much more complex task than had been initially contemplated. Thus, it was necessary to work out ways of transforming the base maps into formats which would allow for every part of the coastal zone to be evaluated as to its aquaculture potential. Effectively, this meant that the factors which control production had first to be identified (these have been termed production variables). Then, within each production variable, it was necessary to work out what aspects of that variable were of varying degrees of importance to shrimp farming. If a mapped representation of the production function could be obtained, then "scores" could be allocated to the various parts of the map according to how important (or unimportant) an area might be. This can best be illustrated by an example. Soils would be a production variable effecting shrimp farming. The characteristics of soil varies over space. For shrimp farms it is important that soils are both non-porous (usually clays) and not too acidic. Some soils will be perfect in both aspects, others may be fairly good, whilst others may be very poor. Scores are allocated to the various areas, ranging from high for good soil combination areas to a low score for bad areas. For some production functions, maps were unavailable. In this case they had to be drawn up using any available data. Annex 2 gave details on all base map construction.
For this exercise it was agreed that, for each production variable, mapped areas would score as 3 for good, 2 for medium, 1 for poor and zero for impossible areas. However, it will be apparent that not all production variables will be of equal importance to any specific form of aquaculture. This means that each variable must also be weighted. In this exercise weightings were allocated such that the raw "scores" for important production variables were multiplied by three, the scores for medium importance production variables were multiplied by two, and the scores for low importance variables were multiplied by one (or left as they were). Effectively, this meant that when a scored map is produced for an important production variable (such as "soils"), then the score ranges on the map were from 0 to 9; whilst on a least important production variable map, such as "access to electricity", then the score ranges were from 0 to 3. A final stage in producing a zoning map for aquaculture could be made by simply overlaying all the scored production variable maps, i.e. within the GIS, superimposing them exactly on top of each other, and then, for all areas of the map, aggregating these weighted scores. A more detailed explanation of this methodology was given as Annex 4 to this consultant's first report.
During the visit all the scoring methods were worked out as they apply to shrimp aquaculture (see Annex 3), and a detailed written and verbal description was given to the PIU of these methods. It should be noted that a few changes have been made from the scoring and weighting descriptions given in Annex 4 of the first mission report. The most important change concerns the flow diagram. Here the combination of production variables has been revised. Details of how the new map layers are to be combined are shown as Annex 4 to this report. It is important to note that here an element of "constraints" has been now incorporated into our scoring, i.e. certain areas can be eliminated from considerations for zoning if it would prove impossible for them to meet certain essential criteria. Other changes include the dropping of the production variable "Hatchery", since it is now clear that distance from a hatchery in Sri Lanka would be no barrier to shrimp farm location, and the use of "Population Density" rather than "Urban Areas", since it was considered that this was more relevant in a country having high levels of rural population. Finally, a variable "Distance to Saline Water" has been added.
During this visit work commenced on the production of the scored maps, and several of these were produced, using the IDRISI GIS. Each was carefully scrutinised to ensure that the mapped output looked appropriate. The exact methods employed in using IDRISI were conveyed by the National GIS Consultant to the other members of the PIU. Given that some of the scoring methods are being newly devised, some problems will occur that will need the consultant's intervention. This will have to be carried out via E-Mail.
The final stage reached during this mission was to practice the requisite overlay procedures, i.e. in order to ensure that appropriate output might be obtained. From the little output produced, it appears that the methods being used are likely to be satisfactory, i.e. within the quality constraints of the input data being used.
Output from IDRISI here was in colour at A4 size. IDRISI works in a raster (pixel) format, and for this exercise the pixel resolution was 50 metres. The scale of the output maps was approximately 1:280,000. It will take a deal of experimentation to achieve map output which is optimised for visualisation, i.e. in terms of colours use and class boundaries defined. Obviously the GIS gives the facility to "zoom" into areas of interest if required. At this stage of the project it is anticipated that, for these areas of interest, numerically ranked scores can be actually provided for individual pixels, thus allowing for a very detailed examination of location potential.
It has not proved possible during this visit to obtain the final zoning maps covering shrimp farming potential in the pilot area. However, if the digitiser were to be delivered this week as promised, and if the remaining scoring methods can soon be applied to the base maps (this should be possible within a few weeks), then this final output could be achieved in about two month's time. Before work then commences on zoning for other types of aquaculture in the pilot area, it will be vital that this author inspects this final output. In order that lines of progress are clear to the PIU, Annex 5 lists a whole series of tasks that the PIU personnel should now be proceeding with, and Annex 6 gives a summarised work plan covering the next three months (April to June, 1998).
When the final output on zoning is achieved for any type of aquaculture, it will be useful to display the results not only in map form but also in tabular form. This means that, for instance, a table could be completed which showed how many hectares of each quality of shrimp farming land was sited in each A.G.A. Division. Obviously data could also be plotted graphically, and statistical analyses could be performed on the final data.
Very little progress was made on this task. There were two main reasons for this. Firstly, it was considered of paramount importance to make as much progress as possible on the coastal aquaculture zoning task. The priorities here were associated with maximising the use of the National GIS Project Consultant, in making certain that the PIU were familiar with all GIS procedures, and in making up for the general lack of progress made thus far. Secondly, after a meeting with the two main organisers of the associated or complementary Australian funded project mentioned in section 1 above (Dr Sena de Silva and Dr Upali Amarasinghe), it was clear that further progress could not be made until some basic decisions on work allocation had been made. Thus, it was agreed at this meeting that the two groups should cooperate, but that it would be necessary for decisions to be made as to exactly "who does what". This will probably be discussed at the next National Project Co-ordinating Committee meeting.
This second mission has proved to be valuable from a number of viewpoints:
6.1
a) The PIU had advanced about as far as was possible within the confines in which they were working. These confines mainly related to equipment shortcomings, and to possibilities for progression on base map construction. The Unit was therefore ready and waiting for more external inputs.b) A large number of minor problems have been "sorted out", and many facets of the project which were in need of some positive impetus were reinvigorated.
c) Some excellent progress was made on achieving the so-called "map scoring" and in obtaining relevant map output.
d) This consultant was able to overlap with the national GIS consultant before his assignment ended, to discuss with him a range of matters, and to ensure that GIS practices had been satisfactorily adopted.
The PIU now appears to well established and is mostly functioning along the lines that were originally intended. To bolster this observation, NARA have just issued their budgetary plans for the 1999 calendar year, and the PIU is to receive funding of 150,000 rupees. This is to cover all operating expenses, excluding salaries and surveys.
However, although positive advancements were made, there are still a range of serious problems or considerations which are in need of attention:
6.2
a) This project has been plagued by slow equipment delivery. As has been described in section 2, there are still items of equipment which have yet to be delivered. The lack of access to a suitable digitiser has meant that there is a considerable backlog of digitising to be done.b) There is still insufficient staffing inputs into the PIU, which is essential if the goals are to be met by the end of the project.
c) It has not proved possible to find adequate data for the construction of all of the base maps, and for certain important parameters, e.g. soil pH, user conflict areas and lagoon salinity; this is a difficult matter to resolve.
d) As a consequence of a) b) and c), the work as a whole is significantly behind schedule with there still being no final output, even for the pilot zone study.
e) Some of the "map scoring" methods are a little complex, and there has been insufficient time to fine tune these.
f) The matter of how this project should integrate its work with the GIS proposals proposed within the University of Kelaniya/Deakin University Project, has to be resolved.
g) A major consideration which must now be faced, is that of the future of the PIU once the TCP project has finished.
The tasks which had been set for this mission have not been fully accomplished. It has not been possible to assemble all of the equipment, but at least the staff have made adequate headway in familiarising themselves with the demands of both the GIS software and the hardware which has been received. Collectively, it would be fair to say that they also have a realistic demand of the requirements necessary to complete the zoning for aquaculture. Having said this, it would be too much to expect them to be able to compile a scoring methodology for each possible type of coastal aquaculture without some considerable help from this consultant. As has been previously shown, not all of the input data has been collected. Some output data was shown to all members of the National Co-ordinating Committee, but at the time of meeting with them, this only consisted of fairly elementary base map data. Obviously, it has not been possible to review the final output of mapping for the pilot area, and reviews of the mapping/scoring methodologies cannot yet be evaluated. Having said this, the output achieved to date looks very promising. Finally, as indicated in section 3, a start has been made on collecting the data and building up the base maps in order to undertake zoning for coastal aquaculture at the national level.
There is a further "conclusive" item which is worthy of mention here. There is an obvious sense in which this whole shrimp farm zoning plan is taking place too late. It is strikingly clear that zoning for this activity should have taken place about a decade ago. This might have prevented the over abundance of farms along the Dutch Canal area. The relevance of mentioning this is that, had the "scoring " been carried out then, the results would probably have been very different. Thus, if all the Dutch Canal farms were excluded from our considerations, then scores for production variables such as "Water Quality", "Distance from Another Farm" and "Floodable Areas" would change - they would be much higher. It might be interesting to carry out the zoning methods described in this report, using a hypothetical "pre-shrimp farming" scenario. This might reveal how good the actual area is (or was not!). It might then be possible to model the expansion of farms in this area, and to arrive at an equilibrium point where it became clear that scores were starting to decline. Obviously some "guesstimation" would be involved, but it seems important to establish a maximum farm loading point for any area, and indeed for each type of coastal aquaculture.
Recommendations
In making these recommendations, the consultant is cogniscent of the fact that some of them might more realistically be made at the very end of the project. However, the project is due to end this coming summer (if it cannot be re-phased), and there are many longer term considerations which need to be made before then. The recommendations are listed in no particular order:
1. The input of working effort into the PIU should be both increased and "formalised" in some way. At present two members of the PIU have other work commitments which prevent them from giving much time to PIU concerns, and one member contributes no time at all. Although promises have been made to increase this time input (up to 80% time input from three members), the staffing problems at NARA might prevent it.
2. As one means to expedite the progress regarding the GIS aspects, it has been suggested that the PIU might take on a full time person with the responsibility for carrying out all the digitising work. With the cautionary note that it is difficult to visualise the amount of digitising work inputs necessary for say more than six months in advance, this seems to be an excellent suggestion. Since one member of the PIU (Dr Siriwardena) is unavailable to take on PIU duties, this "digitising person" would effectively fill the vacant fourth position in the PIU. The consultant recommends that Dr Siriwardena be asked to fulfill a co-ordinating/advisory role to the Unit. This seems realistic in the sense that he is working at NARA full time, he takes an interest in their activities, he is on the National Co-ordinating Committee and he is a director of aquaculture research within NARA.
3. I understand that there is some discretion regarding the temporal limit to the TCP project. If it is obliged to be wound up after 18 months, from the GIS output point of view, there is almost no chance that it will have accomplished the anticipated outputs. The reasons for this have been clearly stated. The consultant therefore recommends a six month re-phasing of the project to allow it to continue until December, 1998. Given this extra time, the work target is more amenable to completion. However, at this point in time it is impossible to predict whether or not all the mapping component can be completed by that date. If the TCP project is extended, careful consideration needs to be given to budget allocations for the PIU.
4. A GIS counterpart expert is not available at NARA or MFARD. Consideration must be given as to who fulfils the role of PIU manager after the TCP finishes. It is recommended that urgent consideration be given to identification of a candidate to take on these duties. In order to give guidance to the Project Co-ordinating Committee, the following is a list of criteria which a new "PIU manager" would need to meet:
a) Should be employed by NARA on a regular basis.b) Should be working full time within the PIU.
c) Should be graded above other PIU members for the additional responsibility.
d) Should be responsible to someone at very senior management level,
i.e. because their duties would cut across Divisional lines
e) Should be very computer literate.
f) Should have a strong sense of "spatial awareness".
g) Should exhibit a strong degree of initiative and inventiveness, i.e. be able to envisage ways in which the PIU could broaden its GIS work.
h) Should be able to engender a strong team identity.
5. The contract of the National GIS consultant (Mr Adikari), is due to end on March 31st, 1998. This will not give sufficient time for him to complete the wide range of tasks required to complete the GIS aspects of the project. The consultant recommends that his services be extended. This consultant has ascertained that he would be available on a part time basis (average of 2 days per week) in the future, and that his employers (Mahaweli Project) would agree to this. A recommendation is made that this arrangement should be entered into for a period of six months. With this arrangement, Mr Adikari's services would be required for about 52 days from the period of commencement. Suggested amended Terms of Reference for this work are given in Annex 7.
6. At the last National Project Co-ordinating Committee meeting, it was agreed that members of the PIU should prepare a collective one month work plan, for submission to the Project Co-ordinator. This should be sent to FAO RAPR, Bangkok, so that activities can be monitored. Annex 5 has set out the work which needs to be carried out in the immediate future, and Annex 6 provides a summarised plan of work covering the next three months in order that the PIU can see priorities.
7. Steps should be taken to ensure that the array of computing hardware remains with the PIU once the project is ended.
8. Copies of all scored maps and overlay maps, plus the final zoning maps, must be sent to this consultant, and to the project coastal aquaculturist consultant, Mr. Charles Angell, for perusal and verification. This might be possible to do as E-Mail attachments if the present NARA modem/communication line facilities can accommodate this. Otherwise, assistance should be requested from FAOR. They could be sent by surface mail as colour print-outs.
9. That some Committee be activated within NARA and/or the Ministry to decide how, or on what basis, the GIS related work should be partitioned between the University of Kelaniya based project, covering the use of GIS to plan for inland fisheries and aquaculture, and the work which was planned to be executed by the PIU under this project. Thus, given that no particular guidance was given as to how this TCP should be involved, and given that NARA/MFARD should have discussed work partitioning when entering into the agreements, then a decision must now be reached. In a meeting with the University of Kelaniya team, including the Australian based project co-ordinator and leader, it was suggested that the PIU contribution might well focus on looking at the relationship between tank distribution and various criteria associated with the markets for fish, and possibly creating a database on the stocking and catches made for specified tanks or reservoirs. The Kelaniya approach was going to be directed towards looking at productivity and production systems within individual tanks. They hoped to build a model which would allow for the prediction of likely output from varying tanks in different agro-ecological zones. This consultant recommends that perhaps the venue for making such a decision should be the next meeting of the Project Co-ordinating Committee, to which representatives of the University of Kelaniya and the Central Environmental Authority should also be invited. In one sense it would be a positive idea to drop the "GIS and inland fisheries" component of this TCP project (seeing that it is way behind the time schedule), but in another sense the PIU should be involved since they must consider work beyond their immediate project tasks. It is recommended that any GIS work associated with the substance mentioned above, i.e. the inland fishery and aquaculture work, be delayed by the PIU until a decision has been reached.
10. The Project Co-ordinating Committee has only met on two occasions since the project commenced, i.e. seven months between meetings. This frequency is not sufficient if the necessary control, support and guidance is to be given to PIU staff. Meetings should be held at three monthly intervals as per provision in the TCP project agreement.
11. There is a perceived need by members of the PIU for further training in the IDRISI GIS. There was no opportunity for this in the course they followed last September at the University of Colombo. Although the IDRISI is a sound product to be using, the associated manuals are almost useless, and the knowledge gained thus far by PIU members has virtually been by trial and error. If funds were available, and someone locally could be found who had a sound knowledge of IDRISI, it would be a good idea to ask them if they would be prepared to conduct a one week intensive course. This could take place at NARA. The consultant recommends that the Project Co-ordinating Committee urgently look into the possibility of arranging this.
12. Once the project period expires, it will be necessary to place the PIU under the auspices of a specific Division within NARA. At the present time it is housed, and rather loosely affiliated to, the Information Division. The consultant recommends that in future it should operate as part of the Aquaculture Division. This is advised since presently all of the work envisaged is connected to aquaculture and three of the PIU members are from this Division. Also Dr Siriwardena (Director of Aquaculture) is familiar with the workings and aims of the PIU. And finally, the leadership within the Information Division are not familiar with the type of work being pursued by the PIU. Having made this recommendation, it is important that the PIU's work is not seen as being confined to the Aquaculture Division. On the contrary, it could and should be positively engaged in a range of aspects of NARA's work.
13. In a similar vein to 12 above, consideration needs to be given as to how the PIU should be linked to MFARD. Being a Unit within NARA, the links with MFARD will be maintained but should perhaps be strengthened through the continuance of the steering committee. The main reason for maintaining the links are concerned with overlap in the resources used (all relating to fisheries in the widest sense), and since MFARD does not have a GIS capability, the PIU could well increasingly perform this function in the future.
14. If the PIU does establish new areas that should be very suitable for any kind of coastal aquaculture, the Project Co-ordinating Committee (PCC) should explore what steps need to be taken to ensure that the suitable areas identified are in fact zoned for aquaculture, and that the findings be extensively promoted and put in practice by all involved entities and committees.
15. That competent persons within NARA and the "about to be established" Aquaculture Authority should, when the first of the final maps have been produced, join together with the PCC to establish a benchmark as to what constitutes "suitability for coastal aquaculture". It is to be determined how good for shrimp farming are in reality the zones (or areas) which we have identified by GIS methods.
16. This consultant, together with the project coastal aquaculture consultant, will be obliged to do a great deal of work (associated mainly with scoring for other types of aquaculture), well in advance of any final mission to this project. This is a result of the very slow progress which has been achieved through lack of PIU productivity, lack of equipment, lack of data, and through the inherent complexity of the task. Of necessity this work will need to be carried out in our respective home bases. At the present time the consultant does not think that the final mission should take up the four weeks which are proposed, though this very much depends upon what is decided with respect to the work on the inland fisheries and aquaculture GIS.
If no activities for inland fisheries are to be carried out, the final mission should be of three weeks duration, allowing one week for home based work by the two international GIS and aquaculturist consultants who would be able to collaborate via e-mail.
ANNEX 1 Itinerary and Persons Met
ANNEX 2 Resources Used and Methods Applied in Construction of Base Maps for Pilot Area
ANNEX 3 Scoring Procedures to be Followed for "Shrimp Farming"
ANNEX 4 Stages in Combining Map Layers to Produce a Shrimp Aqauculture Zoning Map
ANNEX 5 Work to be Carried Out by PIU in the Immediate Future
ANNEX 6 Plan of Work for the PIU over the Three Months
ANNEX 7 Proposed Amended Terms of Reference to Cover an Extension of the National GIS Consultant's Work on TCP/SRL/6712
ANNEX 8 Bibliography
|
7.3.98 |
Depart UK. |
|
8.3.98 |
Arrive Colombo. Prepare for meetings at FAO and MFARD. |
|
9.3.98 |
Briefing at FAO with Mr G. Bernard, FAO Representative, and Mr K. Sugathapala, Programme Assistant, to outline programme during this visit. Meeting with NPD, Mr Upali Pushpakumara, at MFARD, also to discuss programme. Work rest of day at NARA, where progress was reviewed (with all of PIU team) since my last visit. |
|
10.3.98 |
Examine some of the map output produced, for its derivation and accuracy. Discuss with Charles Angell, FAO Shrimp Aquaculture Expert, TCP/SRL/6712 progress since last visit. Revisit Mr Bernard (FAO) to update him on progress which has been made with setting up of GIS equipment, and with the PIU team. |
|
11.3.98 |
Revisit FAO in another attempt to resolve the GIS equipment situation. At NARA, go through base map construction criteria with Ajith and Charles Angell. |
|
12.3.98 |
National "Moon Day" holiday. Work on potential "map scoring" methods, and type up Annex explaining map construction methods. Prepare for meeting with representative from University of Kelaniya. |
|
13.3.98 |
Meeting at NARA with Dr Upali Amarasinghe (University of Kelaniya) and Dr Sena de Silva (Deakin University, Australia), to discuss the nature of an Australian supported project which seeks to help the development of inland fisheries and aquaculture through the use of GIS and remote sensing. More work on describing map derivations. |
|
14.3.98 |
Rest day. Interpolate 5 m contour in pilot area. Search directories for more data sources. Type up report. |
|
15.3.98 |
Rest day. Continue with report. Make out instructions for updating base maps used in pilot survey. Investigate capability of IDRISI to do map scoring. |
|
16.3.98 |
Meeting with Dr Upali Pushpakumara (National Project Co-ordinator) at MFARD to discuss outstanding problems, and agenda for meeting of the PCC. Meeting with PIU members to discuss immediate factors which need pursuing. Meeting with Mr Sugathapala at FAO to settle more equipment problems. |
|
17.3.98 |
Visit EMSO Ltd, to try to speed up the delivery of the digitiser. Work on scoring methods to be used. Attend meeting of the Project Co-ordinating Committee held at MFARD under chairmanship of the Ministry's Secretary. |
|
18.3.98 |
Further work on scoring methods and on final report. |
|
19.3.98 |
Visit to Olaf Haub (UNDP) who is working at establishing a GIS and database for the Resettlement and Rehabilitation Authority of the North. Supervise first output of the "scored maps" for the pilot area. Work on report. |
|
20.3.98 |
Work with Mr. Angell on finalising the base maps and the scoring methods for the pilot area. |
|
21.3.98 |
Rest day. Work on report and review more literature on the pilot area. |
|
22.3.98 |
Rest day. Work on final scoring procedures, and on final report. |
|
23.3.98 |
Work on difficulty of mapping salinity and water quality. Talks with Ms. Barbara Bierhuizen (BOBP, Madras) on progress in our respective projects. |
|
24.3.98 |
Finalise work on map "scoring" proposals. Meeting with Ms. Dora Blessich (FAO, Bangkok), Dr Upali Pushpakumara (MFARD), Mr Sugathapala (FAO, Colombo) and members of PIU to discuss project proposals and problems. Discuss work for the immediate future with PIU members. |
|
25.3.98 |
Debriefing with Mr Bernard at FAO. Completing scoring methods and base mapping instructions. Meeting with PIU members to go over instructions on base and scoring map constructions. Meeting with Mr. Uwe Barg (FAO Fisheries, Rome) on problems of enactment and institutionalising project and recommendations. |
|
26.3.98 |
Depart Colombo 03.00 hours. Arrive Rome 16.00. Work on completing final report. |
|
27.3.98 |
Meet with Dr Kapetsky (Inland Fisheries and Aquaculture, FAO, Rome) for final debriefing session. Return to UK. |
In order that any GIS can be brought into full functionality, it is essential that the system captures a range of mapped data, or data which is capable of being mapped. This annex is directed towards explaining how the necessary range of mapped data, for fulfilling the demands of this TCP, was acquired. Some of the mapping was accomplished using the facilities of the Mahaweli Project - this was necessary since the PIU did not have sufficient access to digitising facilities. All mapped data has been superimposed on a "coastline" map (see below). All maps are held both in a vector format within the GIS Arc/View, and in a raster format within the IDRISI GIS package. The descriptions which follow mainly relate to maps for the pilot area only, though it is assumed that similar resources and methods can be used to make up the maps covering the whole of Sri Lanka. Finally, comments have been included as to the assumed quality of each map, and in some cases suggestions have been made as to how they could be improved.
1. Areas Liable to Flooding
During the monsoon season especially, certain low lying land is prone to flooding. Obviously this could have a severe negative effect on any aquaculture facility. Details on the location of these areas were obtained from the Director of Irrigation at Puttalam, and were then superimposed on a 1:50,000 topographic map for later digitising. Some additional flood prone areas were later added, i.e. those areas shown on the topographic sheets as being classified as "marsh or wetlands", and which were not presently being used for shrimp farming.
This data is of good quality, though under very heavy monsoon rains, it could be surmised that localised flooding would occur, and that areas on the lowest flood plains of most rivers would be inundated.
2. Availability of Ground Water
Many shrimp farmers are obliged to have access to fresh water. This is used to dilute the salinity levels, mainly during the dry season when levels become too high for shrimp culture. Ground water may be used if surface supply from rivers is not available, and the use of ambient sea water can also reduce high salinity levels. Data for this map was obtained and digitised from a map originally published by Fernando (1988), and reproduced in a Wetland Site Report (CEA, 1994). Water quantities are recorded in terms of the Net Annual Ground Water Recharge in mm.
This data should be treated with caution, since maps published by Anon (1990, b,c,d) describe ground water in parts of the pilot area as "extensive and highly productive aquifers", i.e. these are in areas which are shown by Fernando (1988) as having an annual recharge rate of only 120 mm p.a. The Water Resources Board should be contacted for up to date data, or for their advice as to which data is most reliable.
3. Bathymetry
Different types of coastal aquaculture operate within various water depth ranges. For the purposes of this project, aquaculture can be envisaged as taking place either above the mean tidal line (shrimp farming), in shallow water (sea cucumbers, crabs, sea grass, mussels, etc.) or deeper water (sea cages). It was decided to plot the approximate 5 metre and 10 metre bathymetric lines in open, marine water areas (non-lagoon environments), since little local aquaculture would occur at greater depths, and for lagoons areas, the actual bathymetry was mapped. The 5 and 10 metre lines were digitised from appropriate admiralty charts, and the lagoon depths were obtained from a range of publications held by NARA, e.g. CEA (1994). If the 5 metre line is not available for the whole of Sri Lanka, then consideration should be given to obtaining it from satellite imagery.
In most cases this data should be very good, though we are not certain whether the 5 metre bathymetric line can be obtained for all of the Sri Lanka coast. It is also uncertain yet as to the extent of detailed bathymetric data for all of the coastal lagoons. The various Wetland Site Reports do give some data for lagoons.
4. Coastline
A dataset comprising of the complete Sri Lanka coastline was obtained, courtesy of the Mahaweli Project. This had originally been digitised from 1:50,000 topographic sheets. Since much marine aquaculture potential lies in use of the many coastal lagoons, it is important to note that all major lagoons have been included in this dataset, i.e. even those that are presently cut off from direct marine influences. These coastal lagoons are thus differentiated from the so-called "tanks". This data is excellent.
5. Conflicts in Water Resources Use
Certain coastal and lagoon areas are presently intensively used for a variety of purposes, e.g. beach seining, tourism and recreation, lagoon fishing, boat haulage sites, wildlife refuges, etc. It is considered that these areas would be unsuited for aquaculture on social or environmental grounds and conflicts in resource use would almost certainly ensue. Details on the location of these sites were obtained from map evidence, from existing knowledge of lagoons and from field inspections. The sites concerned were mapped via digitised "strips" extending along the selected shoreline in known areas of user conflict.
This data should be treated with caution. Whilst in general it will have a degree of accuracy, no effort has been made to quantify the degree or the likely amount of any user conflict. A survey carried out in 1989/90 (NARA, 1990), provides an excellent summary of the types, and the amount of conflicts which might occur around (or on) many lagoons. It will be important to slowly build up and revise this dataset, mainly through visual evidence collected on field surveys.
6. Existing Shrimp Farms
Details on the whereabouts of existing shrimp farms were obtained initially from records kept by the NW Provincial Council at Chilaw, and other secondary data were obtained from an early map. However, since it is known that many, mainly small, illegal farms exist, it was necessary to carry out field investigations to locate as many shrimp farms as possible. The field investigations were also important in assessing the actual boundaries of the farms. It is still possible that not all farms have been accounted for. It is likely that a better way of locating farms in the future will be to use recent high resolution satellite imagery, but this is not yet available to the project (though it might be available before the project finishes). The location of all farms found was recorded initially onto the topographic base maps, i.e. for later digitising.
This map is quite good, though we are aware of the likely existence of a few more farms. The availability of a promised satellite image may allow for the updating of this map.
7. Hatchery Location
For some types of coastal aquaculture, it is of some importance that hatcheries are within a reasonable distance from the farm. However, for other types of farming, e.g. shrimp farming, fry may be transported over some considerable distance. Although hatcheries for shrimp will not be included as a production variable in this exercise (as had originally been planned), the location of these hatcheries was still mapped. Data on shrimp hatchery location was obtained from the Ministry of Fisheries and Aquatic Resources Development. Since this list of "Approved Shrimp Hatcheries" was somewhat out of date, the present functioning of hatcheries was verified in the field, and their location was accurately recorded using a GPS. It is possible to get production figures from most hatcheries. The location of hatcheries was simply recorded by point symbols on a base map.
This map should be quite accurate, though no indication has been given as to the size of hatcheries. Also, hatcheries which are annexed to farms have not been included, and sometimes these hatcheries may have surplus stock for sale.
8. Low Lying Coastal Land
For land based forms of marine aquaculture, e.g. shrimp farming, it is important that the land used should be of a low elevation. Thus it is impractical to lift water above sea level, in any significant quantity, to more than 5 metres. This low-lying land should be adjacent to saline water. The coastal area which might theoretically be utilised for shrimp farms has been defined in the context of this mapping as being all land areas inwards from the sea (or a lagoon) which lie below the 5 metre contour line, although in practice it is unlikely that farms would wish to be sited at more than about 2 kms from mean tidal level. This contour could be interpolated from 1:50,000 topographic maps, i.e. using the 10 and 20 metre contours as reference lines.
This map is only reasonably accurate. It has not been possible to acquire information to allow the 5 metre contour to be accurately plotted. Information to allow for this should be sought - possibly from the highway authorities.
9. Population Density
The density of population can have effects on aquaculture locations which are similar to those under the heading of "Urban Areas". Thus a positive influence may be as source of farm labour, and negative influences may be that heavily populated districts experience high land costs, pollution and poaching. Generally it is considered that aquaculture in Sri Lanka is more likely to thrive in areas having a low population density. Since the rural population density in Sri Lanka is generally high, it was considered that a production variable showing population density would be more relative to fish farming location than the variable of "Urban Areas", i.e. since there are relatively few coastal towns. There are population density maps available which are based on latest census counts, and are mapped to A.G.A. Division level (now called Divisional Secretary's Divisions - DSD). These were simply digitised.
This map is unlikely to be very accurate because, unfortunately, the census material is very dated - 1981. However, since only relative density was being considered, and this is unlikely to change dramatically, then the use of old data was not considered to be a problem. More up-to-date information should be sought.
10. Power Lines and Transformers
Some electrical supply is important for most forms of shrimp farming. This is primarily for water aeration purposes. Maps were obtained from the Ceylon Electricity Board - Regional Office for the NW Province at Kuliyapitiya showing both the main high voltage power lines and the location of all the transformers. Each transformer has an 11,000 volt supply, and power can be "re-transmitted" from these to surrounding areas. However, for supplies to a shrimp farm, the power would not be transported further than about 3 kilometres from a transformer. It is possible to farm at some distance from an electrical supply by using power generators, but this is both expensive and risky. Both the power lines and the transformer locations were digitised onto a base map.
This map is likely to be accurate, though the source data we have was undated. The most up-to-date information should be sought.
11. Protected Areas
These are classified as areas where, for a number of reasons, aquaculture would not be permitted. The reasons for this include; National Parks, salt pans, government property and establishments, mangroves, etc. The actual sites were simply superimposed onto a coastline base map, using a different colour for each main category.
The information here should be quite reliable, although in practice, due to the current situations throughout north and east Sri Lanka, it is likely that there would be huge tracts of land where effectively shrimp farming enterprises could not be established. A check needs to made on whether there are other types of conservation areas in which aquaculture would be positively disallowed.
12. Rivers
Rivers might be important to coastal aquaculture because they represent a potential source of fresh water which might be needed if salinity levels get too high. Rivers were digitised from 1:50,000 topographic sheets. For this exercise, only wider (larger) rivers have been captured, i.e. those which are shown as double lines on the base maps. This was justified on the basis that smaller streams would be unlikely to provide a reliable source of water, especially during dry seasons. However, it is important to note that there is insufficient stream flow data to corroborate this assumption. The Dutch Canal, linking Chilaw via Mundal Lake to Puttalam Lagoon, and the section going south from Chilaw to Chilaw Lagoon, have been excluded as rivers since they only carry saline water.
Generally, this map will have great locational accuracy, but it will be unuseful if data is required on water volumes or river regimes. Additionally, we are uncertain of the criteria which is used to differentiate between rivers drawn using a single or using double lines on the topographic maps.
13. Road Access
This is clearly important for the transportation of inputs and outputs from the aquaculture facilities. All main and minor roads were digitised from the 1:50,000 topographic sheets (those shown in red or yellow). Cart or jeep tracks were excluded as they were considered to be unsuitable for regular vehicular access, though it is accepted that, were a good site to be found at a distance from a road, then it might be viable to improve road access.
This map should be very accurate, though it may need updating - certainly if it is pre-1990.
14. Soil pH
This production variable is extremely important to shrimp culture, since for a number of reasons, high acidity levels prove extremely detrimental to shrimp production. From the point of view of this exercise however, mapping soil pH is very difficult to do. This relates to the comparative lack of data (NARA has about 75 pH readings for the whole of the pilot area), and to the fact that pH may vary considerably at the micro-scale. In order to partially overcome this problem, a "soil pH trend surface" has been generated. This map was achieved by a plotting of the approximately 75 soil pH (KCl) survey sites as shown in work carried out be Dr Jayasinghe, and then using the INTERPOL function of IDRISI. The use of these 75 sites seems realistic in the sense that they are widespread, random and they were all sites in areas where shrimp farms might well be established. This surface reveals that there is a definite pH gradient which declines from north to south throughout the pilot area.
As an indication of pH levels at particular sites, then this map will be far from reliable, but the surface provides a good general indication of likely pH levels which might be expected.
15. Soil Structure
Soils are a major consideration to shrimp farming from two perspectives, i.e. their acidity (pH value) and their porosity (water retention capability). Data on soil structure was obtained from maps (at 1:63,360 scale) purchased from the Coconut Research Institute (1994), and digitised using the Mahaweli facilities.
This base map appeared to have been drawn originally with great accuracy. Soils have been classed into numerous series. From the perspective of aquaculture, they are not always easy to interpret, i.e. since we require a good measure of their porosity (which basically refers to the proportion of sand, silts and clays in their make up), and no data is given on this. However, inferences could be obtained from the descriptions given in the key.
16. Urban Areas
Urban areas might be important to shrimp farms both as a source of inputs and labour, although they could also be detrimental insofar as land costs are likely to be high and there would be increased potential for pollution and for poaching. Urban areas can easily be identified on the 1:50,000 topographic sheets. The periphery of each urban area was simply digitised.
This map is likely to be accurate though no indication can be obtained of the actual size of the town (urban area).
17. Water Salinity
Salinity is a most important determinant of water suitability for shrimp farming. Optimum range for salinity is 15 - 25 ppt. Above 45 ppt, and below 10 ppt conditions are very poor. Since optimum conditions would seldom be found, shrimp farmers rely on an ability to adjust the salinity range by the addition of more or less saline water from marine or fresh water sources. Obtaining data on salinity levels and variations is difficult, although some does exist. For this exercise it was therefore determined that a "model of salinity" could be formulated. This was constructed by the inclusion of those main factors which actually determine (control) what salinity levels might be. This model would show the likely salinity variations for lagoons throughout Sri Lanka. It would not be applied to open marine water areas - these remain relatively constant throughout the year. Upon advice from the chief oceanographer at NARA, the main factors controlling salinity were assessed as:
a) Surface area of water body. Smaller lagoons are usually shallower, and they dry up quicker giving rise to high salinity levels.b) Annual rainfall. Lagoons in areas of higher rainfall will have reduced salinity levels because of this fresh water input.
c) River inflow. The more fresh water which flows into a lagoon, the lower will be the likely salinity levels.
d) Tidal/exchange rate. Where tidal flushing is high, then there is the propensity to reduce salinity levels. In rare cases the inflow of fresh water will be higher than the tidal flushing rate, in which case flushing may increase the salinity levels.
Obviously other factors will be important, e.g. evaporation rates, but in the main these factors would not vary significantly throughout Sri Lanka. Data for input to this "Model of Salinity" was obtained from the following sources:
a) Surface area of water body. In many cases the size of marine lagoons can be obtained from the various CEA "Wetland Site Reports". If this is not available then lagoon area can be assessed from topographic maps.b) Annual rainfall. This can be obtained from page 39 of the National Atlas of Sri Lanka.
c) River inflow. Actual river discharge for most main rivers is to be found on page 38 of the above atlas.
d) Tidal/water exchange rate. This rate would be a function of the relationship between the volume of water in the lagoon, the tidal range and the cross sectional area of the lagoon inlet. Since these three parameters would be hard to obtain, for the purposes of this exercise an approximation can be derived by calculating the "inlet width" to "surface area" ratio, i.e. using the formula
surface area of lagoon (square metres)/mouth width of lagoon (metres) = ratio
Two examples of doing these calculations are:
(i) For Puttalam Lagoon the surface area is approximately 64 square miles, and the mouth of the lagoon is about 1.25 miles wide. 64 divided by 1.25 equals 51, so the ratio is 1:51.(ii) For Koggala Lagoon the surface area is 727 ha, and the mouth of the lagoon is only about 40 metres wide. 727 ha equals 7,270,000 square metres and this divided by 40 equals 181,750. The ratio here then is 1:181,750.
Instead of constructing a map for this production variable, all calculations were assessed via a score ranges which are described in Annex 3. Obviously, the results of this modelling process can only be crude. Nevertheless, they should give some indication of the likely salinity of lagoons.
18. Water Quality
Water Quality is one of the major variables controlling the location of aquaculture activities. In general terms, the considerations relating to water quality cover chemical and biological determinants such as dissolved chemicals, sewage, BOD, dissolved oxygen levels, agricultural run-off (pesticides, herbicides, fertilisers, etc), suspended sediment loads, industrial discharges, etc. Unfortunately, there is either only very general or sometimes point specific data on water quality in Sri Lanka, i.e. the data that exists usually comprise of "scattered readings for small areas". Following discussions with various NARA personnel and with this project's International Aquaculture Consultant, it was agreed that, in general, an assumption could be made that the marine water quality could be said to be of no problem in Sri Lanka except in the following types of location:
a) Marine waters around urban areas.b) Marine waters around larger river mouths (exits). Data on all of these points can be found on page 39 of the National Atlas of Sri Lanka (Survey Dept, 1988).
c) Marine waters around known larger point sources, e.g. factories or tourist resort developments. Data need to be sought on these.
d) In saline lakes and lagoons which were not directly open to the sea, e.g. Mundel Lake.
e) Along the Dutch Canal between Chilaw and Puttalam Lagoon.
For any of these types of location, the actual degree of the "pollution" problem could not be specifically measured, though as mentioned above, some data exist for some areas. For the purposes of this exercise only larger rivers were included, i.e. those which at least crossed the coastal area being mapped, i.e. they were >10 kms in length. Lagoons were classified as having varying direct access to the sea, according to their "mouth width" to "water area" ratio, i.e. as described in section 17 above.
A map was constructed which recorded these types of locations as either different colour points (for each of urban areas, river exits, and point sources), lines (Dutch Canal) or areas (inland lagoons or lakes). When the mapping is carried out for the whole of Sri Lanka, then two other sources of marine water pollution will need to be incorporated:
a) That of pollution from passing marine oil tankers which empty out their tanks causing tar balls which wash onto beaches. This mainly affects the SW coastal areas, and this would only minimally affect aquaculture.b) Areas of intensive agriculture in the coastal belt. This is mainly the growing of rice and vegetables. Since a range of farm chemicals are frequently used, there must be some run-off into surrounding coastal waters, e.g. on the Kalpitiya peninsular where onions and gherkins are produced (CEA, 1994,a).
It will be clear from the details given above that this map is certain to lack accuracy. It will be essential that every effort is made to try to find actual data on water quality. As a useful indicator of turbidity, and of the direction in which the currents may carry river discharge water, then consideration should be given to using satellite imagery in the future. Unfortunately, because "quality" comprises of several indicators, and because water quality may vary significantly at the micro scale, then it will be a long time before great accuracy can be obtained for this variable.
In allocating "scores" to be used for any type of enterprise, it is important to remember that they must be relevant to certain fundamental criteria which underlay societal values and aspirations. These societal values and aspirations will usually be manifest in the type of political system or party which is in place, and which will be in a position to execute development policies. Within the context of Sri Lanka, we are assuming that developments comply with certain democratic norms that usually apply in a free market economy. So, for instance, we are assuming that any aquaculture development will need to operate on the principle of cost minimisation, i.e. the producer will be seeking to reduce his/her costs as much as possible. This is extremely important to the whole concept of "scoring" because, if the assumption of cost minimisation was not made, then aquaculture could take place almost anywhere.
The actual basis on which scoring and scoring ranges are achieved was described in Annex 4 to this consultant's first mission report. This information was previously given to members of the PIU. What follows are details on specific scoring procedures to be applied to zoning for shrimp farming. The scoring should be applicable to both the pilot area and to the whole of Sri Lanka. Please note that "Access to Hatcheries" has now been dropped from this exercise. Also, a simpler method has been applied for actually combining the various scored variables, i.e. compared to the method described in section 6 of the "Instructions for Aquaculture Zoning Using GIS" (see below). The lower limit of the score range for most of the Production Variables has been changed from the original 0 (zero) to one. This was because it was decided that only a score of zero should be used when conditions were absolutely impossible for shrimp farming, e.g. on Protected Land.
New Method for Producing the Shrimp Farm, Optimum Location Map
These instructions replace those given during the first consultancy in the document "Instructions for Aquaculture Zoning Using GIS" - section 6.
This stage is simply a matter of combining the various scored layers. This is done using the IDRISI GIS. Progressively, two layers are combined, using the OVERLAY function, to create a third (new) layer. The attached sheet shows the order in which this should be done for shrimp farming. As a change from the previous instructions, the scores for one layer will be added to the scores for the other layer. The use of multiplication (rather than addition) was previously chosen because it was suggested that, if some variables received a zero score for any area, then these areas would be eliminated from any consideration for shrimp farming because if a zero is used in multiplication, then any answer (or final score for an area) is automatically zero. However, the same result can be achieved if necessary by screening out areas, i.e. the "Protected Areas" layer has been added to the "Low-Lying Land" layer, and their combined layer will be deducted from the aggregate score total. The use of addition rather than multiplication will result in a more correct procedure, and it will mean that the actual integers being used are not too high.
Each of the layers (coverage) described below will conform to the layers as listed on the attached systems diagram
1. Water Quality
The scoring range for this variable is from 1 to 9. The base map for Water Quality was specifically constructed with aquaculture in mind, i.e. it could not be said to be a universal water quality map. Five types of potentially polluting sources were identified, and were then mapped as lines, points or polygons. It was proposed that each source type would have its own "likely degree of pollution", and for scoring purposes this means that different sized buffers should be drawn around each. Obviously this is a simplification of the real situation, but without the necessary data it was impossible to state how severe the pollution was (or might be) in any particular location.
The sizes of buffers used and the scores to be allocated are as follows:
|
A. Point sources of pollution |
|
|
|
|
0 - 0.25 kms |
score 1 |
|
|
>0.25 - 0.5 kms |
score 3 |
|
|
>0.5 - 0.75 kms, |
score 5 |
|
|
>0.75- 1.0 kms |
score 7 |
|
|
>1.0 kms |
score 9 |
|
B. Urban areas |
|
|
|
|
0 - 0.5 kms |
score 1 |
|
|
>0.5- 1.0 kms |
score 3 |
|
|
>1.0- 1.5 kms |
score 5 |
|
|
>1.5-2.0 kms |
score 7 |
|
|
> 2.0 kms |
score 9 |
|
C. River mouths |
|
|
|
|
0 - 0.5 kms |
score 1 |
|
|
>0.5 - 1.0 kms |
score 2 |
|
|
>1.0-2.0 kms |
score 4 |
|
|
>2.0 - 4.0 kms |
score 6 |
|
|
>4.0 - 8.0 kms |
score 8 |
|
|
>8.0 kms |
score 9 |
|
D. Dutch Canal |
|
|
|
|
0 - 1.0 kms |
score 3 |
|
|
>1.0 kms |
score 9 |
|
E. Lagoon ratios (see separate sheet, Annex 2 (section 17), for explanation) |
||
|
>1:5,000 |
score 1 |
|
|
>1:1000 - 1:5000 |
score 3 |
|
|
>1:200- .1:1000 |
score 5 |
|
|
>1:50- 1:200 |
score 7 |
|
|
<1:50 |
score 9 |
|
A 1.0 km wide buffer would need to be drawn around the lagoon, and land within this area would be scored according to the lagoon ratio score above.
2. Water Salinity
The theoretical score range for this variable could be from 3 to 12. Going beyond the normal maximum range of 0 to 9 is justified since, not only is this variable so important to shrimp farming, but also any scoring system which uses four factors which must be combined to make up a final score, makes it difficult to use the normal range. Since the general ideal for shrimp farming would be to achieve a salinity which was considerably lower than the ambient sea water salinity, then the scoring is designed to favour areas or factors which lead to lowest salinity values. Data for rainfall and river inflow come from pages 38/39 of National Atlas of Sri Lanka. Each of the four factors being used were scored as follows:
|
a) Annual rainfall: |
Less than 1000 mm |
score 1 |
|
|
1000 to 2000 mm |
score 2 |
|
|
More than 2000 mm |
score 3 |
|
b) River inflow: |
|
|
|
|
Inflow less than 25 cu.m x 10 |
score 0 |
|
|
Inflow > 25 to 100 cu.m x 106 |
score 1 |
|
|
Inflow >100 to 500 cu.m x 106 |
score 2 |
|
|
Inflow > 500 cu.m x 106 |
score 3 |
|
c) Lagoon surface area: |
|
|
|
|
Area less than 200 has |
score 1 |
|
|
Area >200 to 2000 has |
score 2 |
|
|
Area >2000 has |
score 3 |
|
d) Surface area/Mouth width ratio: |
|
|
|
|
Ratio greater the 1:25,000 |
score 1 |
|
|
Ratio 1:25,000 to 1:500 |
score 2 |
|
|
Ratio less than 1:5 00 |
score 3 |
Then a table must be produced which lists all the marine lagoons around the coast. Coastal inlets which have a very wide opening to the sea can be excluded, e.g. the main part of Trincomalee harbour; lagoons which are totally cut off from the sea should be excluded such as those with blocked entrances/field verification required. The pilot area encompasses only Mundel Lake and Puttalam Lagoon. The table will need headings showing each of the above four factors, plus a Score Total column, e.g.
|
Name of Lagoon |
Rainfall |
River inflow |
Surface area |
Ratio |
Total |
|
|
Score |
Score |
Score |
Score |
Score |
|
Mundel |
|
|
|
|
|
|
Puttalam |
|
|
|
|
|
The table is to be simply compiled using the sources described in Annex 2, or by calculating the ratios and areas, and then reading the score from the above score ranges. Some of the "surface area/mouth width" ratios might require change because it was difficult to calculate what these might be without spending a large amount of time finding the dimensions.
A scored map needs to be constructed. The existing coastline map can be used as the base for this. All non lagoon (sea) areas should be allocated a score of 6. Lagoons should be allocated scores obtained from the above table. A buffer of one kilometre should also be drawn around each lagoon, and this should be allocated the same score as the lagoon. All other areas of the map should be scored as 1, i.e. indicating that salinity would be very poor (in fact not even a consideration).
It will be absolutely essential to verify the results of this work by checking against any existing lagoon salinity, and by discussing the results with NARA Oceanography Department. Remember that the results will only be giving an idea on overall salinity - they will not be showing how suitable the salinity is specifically for shrimp farming. One factor that might want to be adjusted is "length of lagoon", i.e. it has been suggested by the project coastal aquaculturist that in long lagoons the water gets more saline towards the end of them (furthest from the sea).
3. Fresh Water
The score range for this variable has been reduced from the original decision - it now varies only from 1 to 3. Fresh water is used to dilute highly saline water, mainly in the dry season. After discussions with various experts, and from information obtained in CEA (1994,b), it was decided not to use the base map showing "Availability of Ground Water". This is because, although some ground water is presently used, this practice is likely to be short-lasting in the pilot area, i.e. since the annual recharge rates are relatively very low and therefore dependable supplies will rapidly run out. High salinity problems may be reduced by simply pumping in ambient open sea water, or via the use of brackish water. Obviously places at a distance from the sea, e.g. along the Dutch Canal, could not get water from these sources. So, for scoring for fresh water availability, it was decided to only rely on rivers as a source. Scoring for this map was achieved by drawing the following buffers around all rivers shown on the "Rivers" base map:
|
0 - 0.5 kms |
score 3 |
|
>0.5 - 1.5 kms |
score 2 |
|
>1.5 kms |
score 1 |
Obviously, in drawing these buffers no allowance was made for either periodicity of river flow, or for the actual amount of water which any individual river could provide. But since only larger rivers were digitised, this should not be too much of a problem.
4. Floodable Areas
The scoring range for this variable is 1 to 3. Although farms must be sited at a low elevation, in order to gain direct access to saline water, they must be cogniscent of the possibility of flooding. This is mainly a problem associated with river flooding during the monsoon season, i.e. it occurs along the lower flood plain of rivers. Although dykes can be built, these would have to be very high if they were to be a secure defence against the largest floods. Buffers need to be drawn around existing flood areas. Scoring would then be:
|
Existing flood areas |
score 1 |
|
|
|
0 - 0.25 kms |
score 2 |
|
|
> 0.25 kms |
score 3 |
5. Soils
The scoring range for this variable is from 1 to 9. Using the base map of "Soil Structure", each of the numerous soil classifications shown in the key was given a rating of 1, 2 or 3 according to its inferred porosity (drainage) capabilities, i.e. 1 being well drained, 3 being poorly drained and 2 being in between. There is no way to gauge the accuracy of this judgement. The digital map was then reclassified to produce a map having only three soils categories - corresponding to their drainage capability. In order to give a range of scores from 1 to 6 (for structure), scores were adjusted such that areas classified as 1 remained as 1, areas scoring 2 were raised to a score of 4, and areas scoring 3 were raised to a score of 6.
Then the "Soil pH Trend Surface" was used. This map was reclassified so that all areas having a pH trend value of greater than 6 scored 3, areas having a pH trend value of between 5 and 6 scored 2, areas having pH values of between 4 and 5 scored 1, and areas having a pH value of less than 4 scored zero. A final "Soil" score was achieved by overlaying the two soil-scored maps, and adding their scores together. This would give a final range of 1 to 9.
6. Road Access
The scoring range for this variable is from 1 to 6. For aquaculture purposes, both main and minor roads can be considered as the same. Scores will vary so that they decrease with increasing distance from a road at an exponential rate. This is because road construction costs would vary approximately exponentially with distance. Buffers need to be created around roads, using the following distances and scores:
|
0 - 0.5 kms |
score 6 |
|
>0.5 - 1 kms |
score 5 |
|
>1 - 2 kms |
score 4 |
|
>2 - 4 kms |
score 3 |
|
>4 - 8 kms |
score 2 |
|
>8 kms |
score 1 |
7. Access to Electricity
The scoring range for this variable is from 1 to 3. Scores will vary so that they decrease with increasing distance from the main electricity supply lines. Costs of accessing electricity would also vary exponentially with distance. Most costs to the fish farmer would be associated with actually gaining access to the transmissions lines via a transformer, and extra costs associated with the length of the supply connection lines would not be so much, i.e. the cost increases would decrease with distance. With distances up to 3 kms, then fish farmers might gain access to an existing transformer. Buffers need to be created around transmission lines, using the following distances and scores:
|
0 - 3 kms |
score 3 |
|
>3 - 6 kms |
score 2 |
|
>6 kms |
score 1 |
8. Distance from Other Shrimp Farms
The scoring range for this variable is from 1 to 6. Although there may be some advantages in shrimp farms being in close proximity to each other, e.g. opportunities for co-operative working, shared equipment, bulk buying of feeds, transference of information, etc., in practice major problems have been experienced in areas where farms are closely clustered. The major problems are associated with water quality and with disease transference. For scoring purposes the hypothesis was made that ideally shrimp farms would be at least 2 kms apart. With increasing distance towards the centre of an existing farm (or cluster of farms), then scores would decrease. Scoring would be as follows:
Existing areas shown as shrimp farms score 1 Buffers around existing farms are drawn and scored:
|
0 - 0.5 kms |
score 2 |
|
>0.5 - 1 kms |
score 3 |
|
>1 - 1.5 kms |
score 4 |
|
>1.5 - 2 kms |
score 5 |
|
>2 kms |
score 6 |
9. Distance from Saline Water
The scoring range for this variable is 1 to 9. This variable had been inadvertently omitted from previous production function considerations. It was considered that any shrimp farm would benefit from having close proximity to its rearing milieu, i.e. it would be difficult and expensive to have to transport large quantities of saline water over any distance. It was considered that shrimp farmers would not normally wish to transport water further than about 1 km. For scoring purposes, the Coastline map was used, and the Dutch Canal was shown as being saline water. Buffers were drawn around the coastline, saline lagoons and the Dutch Canal. Scoring was as follows:
|
0 - 0.25 kms |
score 9 |
|
>0.25 - 0.50 kms |
score 7 |
|
>0.50 - 0.75 kms |
score 5 |
|
>0.75 - 1.00 kms |
score 3 |
|
>1.0 kms |
score 1 |
10. Conflicts in Water Resource Uses
The scoring range for this variable is 1 to 9. It was generally considered that coastal sites which had existing, long standing usage would prove detrimental for coastal aquaculture practices, i.e. since the use of these areas would lead to conflicts over resource use, as well as to loss of land, water rights and probable employment (or income). Conversely, sites which had very little existing use, or which were further away from existing high intensity use areas, should make fish farming sites. Detrimental sites can be envisaged as being either coastal margins (usually beach sites plus the adjacent inshore water), or margins of inland water bodies, e.g. lagoons and lakes. All sites concerned have been digitised as narrow strips along the shoreline, and categories of usage are indicated in the key. For all of these sites, the penalties of using the actual sites for shrimp farming would be very severe, but this penalty effect would rapidly decrease over just a short distance from the site. Buffers were drawn around all sites as follows:
|
0 - 0.25 kms |
score 1 |
|
>0.25 - 0.5 kms |
score 3 |
|
>0.5 - 0.75 kms |
score 5 |
|
>0.75 - 1 kms |
score 7 |
|
>1.0 km |
score 9 |
11. Population Density
The scoring range for this variable is 1 to 3. It was generally considered that areas having high population densities would have relatively adverse conditions for shrimp farming. Obviously there are no objective measurements to give the degree to which this applies. Scoring for this variable was as follows:
|
A.G.A. Divisions with a density of >1250 per sq mile |
score 1 |
|
A.G.A. Divisions with a density of 250 to 1250 p.s.m. |
score 2 |
|
A.G.A. divisions with a density of <250 p.s.m. |
score 3. |
12. Land Above 5 Metres
Scores for this variable are assigned as a zero for all land over 5 metres above mean sea level, and a one for areas below the 5 metre contour line. This reflects the fact that land above 5 metres would be impossible to use for shrimp farming.
13. Protected Areas
Scores for protected areas are assigned as a zero for any protected area, and a one for all other areas. This reflects the fact that shrimp farming would be prohibited in those protected areas shown on the map. This map will be overlaid with the "Land Above 5 Metres" map, and then the resulting "Constraints" map will first be reclassified so that all areas showing a score of 2 will be changed to one, and then it will be multiplied by the "Production Variables Total" map. The result of these procedures will be to produce a final map which screens out any areas where constraints occur.
Following this consultant's visit, it is now possible to outline a range of tasks which will need to be undertaken over the next few months. In the final version of this list, work will be prioritised by fitting it all into a work plan form. In no particular order these are:
1. It will still be necessary to gather more published (or other) data, i.e. in order that the final maps are as accurately produced as possible, and perhaps to establish some new databases. Possible sources still to check include
a) Possible useful sources from the 1996 Directory of Biodiversity. (The consultant provided a detailed separate list.)b) All the sources identified last year (in a note dated 13th August + Annex 3 of first mission report)
c) Dr Robert Riethmueller of the Sri Lankan-German regional Rural Development Project, 109 Yatinuwara Street, Kandy, might have some databases which could help.
d) The Wetland Site Reports, published by the CEA, should prove extremely useful when starting on the full Sri Lanka zoning. Also useful here will be the two volumes published by NARA (1990) called "Survey to Identify Suitable Areas in the Coastal Belt of Sri Lanka for Prawn Culture". The CEA reports offer extensive bibliographies which should be consulted where possible.
e) Mr Olaf Haub at the Resettlement and Rehabilitation Authority for the North, 16/3 Cambridge Terrace, Colombo 7 is doing some GIS work on the Jaffna peninsular under a UNDP project. Although he does not have much useful data (because the project has only just started), he informs that Thomas Leichtle (of the same address) is presently visiting Jaffna to gather data on fresh water supplies on the peninsular - mostly underground supplies. He will be back in Colombo between 21. 3. 98 and about 28.3.98, and will be coming back to Sri Lanka later. He should be able to tell all about fresh water resources (especially underground) in the Jaffna area. He can also be contacted at [email protected].
f) Mr. C. Angell has suggested that the Highways (Roads) departments (Ministry) might have data which allows for the mapping of the 5 metre contour. When they do the surveys for laying roads they must collect this sort of data.
g) Find out whether any of the coastal areas have official designations such as "conservation areas". The NARA (1990) report mentioned in d) above is useful.
h) An effort was previously made to gather borehole, groundwater data, but apparently it would have been too expensive to obtain. This will now be available free of charge, should be obtained. Both the Water Resources Board and the National Water Supply and Drainage Board (Galle Road, Ratmalana) are likely to have a range of groundwater data. Although ground water is now not included as a consideration in the pilot area, it will probably be important in other parts of Sri Lanka.
i) Since one of the forms of aquaculture to be considered is fin-fish culture in cages, wind direction and strength might be an important consideration. Data on this should be obtained from the Meteorology Ministry.
j) If population density data exist for any coastal areas, down to the smallest divisional level, i.e. the Grama Nilhadari Divisions (GND), it should be plotted.
k) It will be very important to seek out bathymetric data for all the lagoons in Sri Lanka.
2. It will also be important to gather some data from field visits, especially from lagoon areas. It must be emphasized that for most forms of aquaculture to be examined (for zoning) lagoon locations will be important. All data gathered in the field must be geo-referenced. The consultant made available separate instruction on how this data should be mapped. Examples of this extra data needed would include:
a) establishing more exact maps showing "Conflict of water resource use", especially of lagoons: Again NARA (1990) gives some good clues on how this might be done, e.g. page 198 of "Phase-2" volume.b) gaining more data on water quality, water salinity or soil pH values: it might be necessary to obtain by field observation. Another factor here which can best be gained from the field is the location of coastal factories or industries, plus holiday/tourist developments.
3. In data gathering, one should consider establishing databases, i.e. a simple form to record the data. A computerised data collection system could be loaded into the PIU laptop. Alternatively, databases can be based on spreadsheets such as Excel. This type of recording is useful when having rows and columns of data. It will be very important to set up databases on factors concerned with water quality, with salinity and with soil pH values.
4. Obviously, all PIU staff should make themselves familiar with the digitiser when it eventually arrives. There are a range of things to digitise. These include both additions to the maps already drawn and some new maps (see 5 below).
5. The following is a list of things which need to be done in order that the base maps for the pilot area are complete:
a) The 10 metre bathymetric line needs digitising, and the bathymetry is needed for the lagoons (Puttalam, Dutch Bay) and for Mundel Lake.b) Extra areas where user conflicts might occur need digitising. The consultants shall advise further on this.
c) The extra areas which are liable to flooding need digitising, i.e. areas shown as marsh or mangrove.
d) The Wetland Site Report (CEA, 1994) indicates that there are a number of shrimp farms shown on the delta of the Mi Oya river, i.e. on the eastern side of Puttalam Lagoon. This needs to be verified and, if they do exist, they have to be marked them in on the map of shrimp farms.
e) Make a new digitised map, called Low-Lying Coastal Land, showing just the 5 metre contour line (The consultant has drawn it in pencil on the 4 topographic (1:50,000) maps).
f) Check that "Protected Areas" map has been merged with the "Mangroves" map.
g) A "Soil pH Trend Surface" map needs to be compiled by using the approximately 75 survey readings made by Dr Jayasinghe, and then using the INTERPOL function of IDRISI.
h) A "salinity" map also needs to be drawn. Instructions for this are given in Annex's 2 and 3. The final map will be a scored map rather than a base map.
i) A Water Quality map needs to be completed. Although I have given five types of pollution sources for this map are available all the point sources of pollution, e.g. factories and tourist facilities, need to be identified. Details on how to complete this map are again given in Annex's 2 and 3.
6) Nearly all the "Scored Maps" need to be completed. The only ones which have been finished are:
a) Road Access.
b) Electricity.The maps needed for this task are shown on the flow diagram - "Stages in Combining Map Layers to Produce a Shrimp Aquaculture Zoning Map". (Annex 4)
7. Organise the installation of the air conditioning unit and follow-up on missing equipment. This is particularly important for:
a) Modem installation in the laptop computer
b) The digitiser.
c) The GPS.
d) The internal card for the Tape Back-up drive.
e) The UPS. Also, arrange for the purchase of the three way switch and the map drawer.8. Each month a one page "Plan of Work for the PIU" should be prepared, and given to Mr. Upali Pushpakumara (Project Co-ordinator) together with plan of field visits and places where data is to be collected with an indication of the present mapping being prepared. The immediate plan in the coming month must include the completion of the digitising for the pilot area (list maps) and the production of the scored maps for the pilot area.
9. It will be absolutely essential that a copy of the scored maps, and the various overlaid maps produced be sent to this consultant as soon as they are produced. He will need to study them carefully to see how good (or reliable) they are! It should be possible to send them as E-Mail attachments, but if this proves complex, then they should be sent surface mail, perhaps on diskette.
10. When good areas for shrimp farming have been identified, it will be essential that field visits are made to these locations. This should be coordinated with the Director of the Aquaculture Division who is familiar with the demands for shrimp farming. The purpose of the visits will be to ensure that the scoring is revealing appropriate areas. If the mapping proves useless, it must be reported to this consultant.
11. A lot of experimentation is needed with the use of colours (or colour combinations) on maps and with classifications, i.e. where best to use class boundaries, and the number of classes to be used.
12. Thinking ahead, you should still be contacting your colleagues to find out what data they hold, and very importantly, what sort of GIS work will you be able to do for them.
13. At the present time, there is relatively little data in digital format. This amount will grow. It is therefore very important that one of the staff sets up a meta-database. This is literally a database to record all data sets held.
14. Mr Adikari, the national GIS consultant, should ensure that all the PIU staff are familiar with the necessary GIS mapping procedures for obtaining the required output. This will mainly be using the digitiser, converting the base maps to scored layers, and to doing the necessary overlay procedures. Most of this final work will be done using IDRISI.
15. If there are IDRISI operations which cannot be worked out, E-Mail IDRISI about this - they are very good about replying. Check that you have technical cover for one year - Mr Sugathapala could advise. If there are uncertainties about the wording of messages, they can be addressed to the consultant to enter corrections as necessary and then forward to IDRISI.
16. Once the back-up drive is received, ensure that copies are made of all files. These need to carefully organised so that they are easily retrieved.
April to June. 1998
The idea of this brief document is to give some idea of the priorities for the PIU regarding the GIS work over the coming three months. The work is to be carried out collectively by the members, although obviously individuals may choose to concentrate on different items. This plan could form the basis of the first three work plans which should be submitted to the NPD, Mr Pushpakumara. These plans are made on the presumption that the digitiser has been received. Also, all instructions refer to shrimp farming zoning unless otherwise stated.
April
1. Become familiar with the digitiser and commence the digitising.2. Complete the field surveys to allow for the User Conflict map to be finished, and to allow for plotting of Point Pollution sources in the pilot area.
3. Collect and compile any secondary water salinity data for all parts of Sri Lanka.
4. Obtain additional bathymetric data for all lagoons, and groundwater data for all parts of Sri Lanka.
5. Do the field survey for shrimp farms to the east of Puttalam Lagoon.
May
1. Do the calculations necessary to make up the salinity tables for both the pilot area and for the whole of Sri Lanka.2. Verify the salinity scoring methods used, i.e. using any data obtained in No.3 for April.
3. Complete the digitising for all base maps in the pilot area.
4. Start setting up any meta databases, and think about how to organise all the IDRISI files.
5. Finish most of the scoring for the pilot area to produce the "scored" maps.
6. Collect meteorological data on wind strength and direction.
June
1. Complete all scored maps in the pilot area.2. Do the overlaying of maps with maps in the pilot area so as to produce a final map.
3. Send copies of the overlay maps to both this consultant and to the consultant, Mr. Charles Angell.
4. Establish methods for collecting data, and set up a database system, covering point pollution data and user conflict data.
5. Visit the areas shown as being good for shrimp farming, and verify them in the field. Pass these results on to this consultant.
6. Experiment with colour combinations for all final mapping (overlaid maps).
Since there have been severe delays in the GIS mapping component of this TCP, a recommendation has been put forward by this consultant that the National GIS Consultant (Mr Adikari) should be asked to extend his period of service to the project. It is recommended that he serves another period of six months, but on a part time basis comprising of two days (to be agreed and formalised) per week. This would amount to 52 days work in total.
The following amended TOR are recommended. All of these tasks are additional to the TOR's which are set out in the initial Project Document. Most of the tasks are of an advisory, rather than practical nature.
1. Ensure that the PIU staff are competent in the use of all GIS hardware items, and that members are familiar with the software upon which they are specialising.2. Assist in the task of training the person NARA is to assign to carry out future digitising requirements of the PIU.
3. Set up a meta database system so that all digital files can be logically stored, easily retrieved and readily inspected, edited, etc.
4. Advise on the collection of additional field data necessary to complete the zoning for shrimp aquaculture in both the pilot area and the national area.
5. Oversee completion of the base maps initially needed for zoning for shrimp farming in the pilot area, and subsequently the base maps needed for the national survey.
6. Oversee completion of all the "scored maps" necessary for zoning for shrimp aquaculture in both the pilot and national areas.
7. Advise on the interests of the PIU at meetings of the National Project Coordinating Committee meetings.
8. Ensure that a system of monthly work plans is instigated, and that plans are submitted to the National Project Co-ordinator and FAO.
9. Advise on any base map construction which may be associated with zoning for all "non shrimp'' forms of coastal aquaculture.
Anon (1990b) Resources and Economic Activities. Wanathavillu Divisional Secretary's Division, Puttalam District. Prepared by Integrated Rural Development Project, Kachchri, Puttalam. pp 107.
Anon (1990c) Resources and Economic Activities. Puttalam Divisional Secretary's Division, Puttalam District. Prepared by Integrated Rural Development Project, Puttalam. pp 139
Anon (1990d) Resources and Economic Activities. Kalpitiya Divisional Secretary's Division, Puttalam District. Prepared by Integrated Rural Development Project, Kachchri, Puttalam. pp 134.
Central Environmental Authority (CEA) (1994,a) Wetland Site Report & Conservation Management Plan: Puttalam Lagoon, Dutch Bay and Portugal Bay. Min. of Environment and Parliamentary Affairs, Colombo.
Central Environmental Authority (CEA) (1994,b) Wetland Site Report & Conservation Management Plan: Mundel Lake & Puttalam Corridor Channel. Ministry of Transport, Environment and Women's Affairs, Colombo.
Fernando, A.D.N. (1988) Journal of the Geological Soc. of Sri Lanka. Vol.1 pp 93-98.
NARA (1990) Survey to Identify Suitable Areas in the Coastal Belt of Sri Lanka for Prawn Culture - Phase 2. NARA, Crow Island.
Survey Dept. (1988) National Atlas of Sri Lanka. Survey Department, Colombo.
