Useful information for assessing forest change can be found in different reports or in only one report providing both sets of statistics. The “FRA method” has been designed to summarise and provide a first screening of the reliable and available sources for forest change. The table below provides a description of all the parameters that are taken into account during the evaluation of information to be used in a change assessment.
Type of change information/data:
This is a descriptive parameter indicating the compatibility of the sources available for estimating change. Four general categories are distinguished:
1. Change information that may be derived from a series or cycle of national or continuous forest inventories;
2. Change information based on two independent inventories;
3. Information from a single specific forest change survey/study;
4. Only state information available.
Coverage:
This parameter indicates the compatibility of the two sources in terms of their geographic coverage. For example, while two sources can be compatible technically they may relate to different areas of land within a given country. Therefore, a source can then be classified as:
1. Covering the whole country, national (N);
2. Only partially covering the country (P), e.g. covering state(s) or provinces.
Periodicity of data:
The respective dates of the two (or more) sources will provide the period for which a change assessment is possible. However, as national inventories generally take place over a long period of time, frequently two or three years (sometimes longer), it is usually necessary to indicate an “average” date for each source utilised in the time series.
Compatibility of time series:
Compatibility of data sources in a time series is defined by three distinct elements:
A: Compatibility of classification schemes used in different inventories:
A1 = High = Same classification has been used in the two inventories;
A2 = Medium = Differences in classification exist, yet still allow a comparison of the two sources (for instance by aggregation or reclassification or compatibility between certain classes);
A3 = Low =Incomparable classifications.
B: Compatibility of methods
B1 = High = Consistent for all inventories;
B2 = Medium = Methodology differs in some aspects between surveys. For example, different remote sensing data used or different field sampling design;
B3 = Low = Methodologies are incompatible and differ in most aspects.
C: Interdependence of time series
Time series is interdependent when interpretation or analysis of the new remote sensing or field plot data is done with reference to the former, specifically to ensure compatibility between inventories
Type of data:
Here the type of remote sensing data (aerial photos, Landsat TM, etc.) should be listed. In general, the higher the spatial resolution, the greater the utility for change assessment
Inventory Field Work:
It helps in evaluating the sampling methodology. The parameter takes in account four categories of fieldwork:
A) Repeated measurements on permanent sample plots
B) One time measurement
C) Combination of the above two
D) None
Remote Sensing Validation
Many countries now rely solely on remote sensing techniques to generate forest cover information. While it is widely accepted that field checks of imagery interpretations to calibrate classifications and polygon delineations improve the overall work, this is frequently not done. This parameter takes into consideration three categories:
A) Statistical field checks (H)
B) Non-statistical (windshield) surveys (M)
C) No field work (L)
Compatibility with FAO classification
It helps to define the compatibility of the national definitions of forest and deforestation (applied in the time series or change study) with FAO’s definitions (see FRA 2000 Working Paper 1, Terms and definitions).
A) High (H): No (basic) differences;
B) Medium (M): Differences in classifications can be overcome through matching/reclassification procedures;
C) Low (L): Major differences between the classifications
Overall quality
Based on all the criteria previously defined, a judgement can be made on the overall quality of the forest change information for estimating national forest change in FRA 2000. Three classes are distinguished: high (H), medium (M) and low (L). In the overall judgement, one should put relatively high value on the compatibility of the time series.
