Annex A: The Canada USA Case Study – Implementation of the FAO Global Ecofloristic Zone Mapping Concept

Regional Ecological Zone (EZ) map and related forest area statistics have been prepared for the North America region, covering Canada and USA (including Alaska, but not Hawaii). Results of this case study illustrate the overall concept, methods, and utility of the proposed classification system, and provide guidelines to develop regional maps for other parts of the world. Source data for Canada were derived from a Canada ecoregion map produced by Environment Canada and incorporated into the Commission for Environmental Cooperation Ecological Regions of North America Map (1997). For the USA, Bailey’s ecoregions map (1998) was used.

In this Annex, the methodology and procedures are presented, followed by results, tables and map, of the case study, including: 1) an ‘equivalence’ table listing the three levels of global EZ and corresponding regional Ecoregions or vegetation types (indicated by labels or codes of input source maps, Table B-1); 2) graphics of the regional EZ at the three levels, and 3) breakdown tables, presenting forest cover by ecological zone and by sub-national unit.

The two breakdown tables illustrate an important use of the FAO EZ maps in the production of forest statistics by ecological zone and by sub-national unit (see below Table B-2 for USA, Table B-3 for Canada). To produce the two breakdown tables for each country, four data items are needed: 1) the regional FAO EZ map, 2) national and sub-national administrative boundary file, 3) an actual forest distribution map, and 4) the most current country-survey data including forest areas by sub-national unit.

The purpose of the two breakdown tables is to provide forest cover estimates by ecological zone, through establishing a spatially explicit relationship between the forest cover areas derived from remote-sensing based forest cover maps and the most current country-survey data presented by FAO. The first table provides a breakdown of forest area according to sub-national unit and ecological zone, derived from overlaying the forest cover map with the country’s administrative boundary map. Total area and percentage (probability) of forest cover are calculated and presented for each ecological zone by sub-national unit (Table 1). Multiplying the percentages in table 1 by the most current forest cover statistics (in column "Total" of table 2) gives the estimated distribution of forest by ecological zone and sub-national unit (Table 2).

The general process for producing the two breakdown tables is as follows.

• Prepare the input data. The EZ map, the country administrative boundary file, and the actual forest distribution map can be either vector or raster format. For the North America example, both Arc/Info and ERDAS Imagine software systems were used to prepare these files and compute the results.

• For the preparation of the EZ map, an equivalence table needs to be developed, similar to that given in Table B-1 for the USA and Canada. The original classes given in the source data (e.g. ecoregions, ecofloristic zones, (potential) vegetation types) are reclassified or recoded to correspond with the FAO EZ Categories as outlined in Table B-1. For Example, the CEC’s ecoregion code ‘5.1: Softwood Shield’ is recoded to FAO EZ class ‘E1-Boreal Plain’, based on the descriptions (CEC 1997) that the ecoregion has a mean annual temperature of –2 to -6C, a mean annual precipitation of 550-1500mm, a vegetation of primarily conifers, lichens and shrubs and that landform is hilly with some plains.

• The country administrative boundary file can be extracted from the Arc/Info Arc-World vector database. So far, there is no better or more updated boundary map than Arc-World. For the actual forest cover map, we should use, in so far as possible, nationally developed actual land cover or forest cover maps from remotely sensed data. For example, the forest cover map for Canada is from the Canada Center for Remote Sensing 1995 1-km land cover map. For the USA, it is the 1993 U.S. Forest Service 1-km forest type map. In the case when an actual forest cover map is not available, or is not suitable, then a substitution should be found. For Europe, we propose to use Joint Research Center FIRS forest cover map. For other regions, the USGS global land cover database may be used.

• FAO will provide the most current country forest area statistics

• The actual land cover/forest cover classification should be re-mapped such that forest land equals to 1 and all other land cover equals to 0.

• Once all input data files are prepared, the computation is done by overlaying the three thematic layers: EZ, country/sub-national boundaries, and the actual forest cover. Calculate the area of forest cover in each unique EZ and political unit combination. Most GIS and image processing software systems have utilities to achieve this.

• Results of the above GIS overlay are then entered into a spreadsheet file, such as Microsoft Excel. Follow examples of the two breakdown tables for each country in this Annex.

References:

Table B-1. Working table for linking the Canada and US ecoregions maps to FAO EZ

Table 2: continued