 |
||
|
Regional Vulnerability Assessment (ReVA) Program |  |
 |
||
|
|
|
|
|||||||||
|
|
ReVA Methods and Data
ReVA began with a simple premise: what we actually know about an ecological system is contained in the scientific data that has been collected on that system. It was clear that data had been gathered by many separate state and federal agencies, different groups within USEPA, and private organizations such as The Nature Conservancy. There was also a major new source of data in satellite imagery of landcover that was beginning to be explored by the scientific community. The task was to collect all of the available information and integrate it into useable assessments of ecological condition. At the same time, experience with large-scale problems such as Acid Precipitation and Global Climate Change indicated the need for an assessment at a larger scale than had previously been done. To address this need, ReVA was structured to deal with regional scale (i.e. multiple states) data and assessments. At these larger scales, it is the spatial pattern of environmental quality that becomes a matter of concern. Having chosen the Mid-Atlantic region for its pilot study, ReVA gathered all available environmental data. The only requisite was that the data had to be spatially distributed across the entire region. This requirement was easily met for landcover data but was more stringent for other sources that had information for only a few sites within the region. For some information, such as air pollutants, spatial coverage was achieved by using state-of-the-art models. For the pilot study, the watershed (8-digit Hydrologic Unit Code or HUC) was chosen as the spatial unit for all analyses. This is a logical unit for assessing environmental quality as there are defined boundaries that are available across the region. It would also be possible to deal with the data using the county as the spatial unit if the endpoint of an assessment was socioeconomic in nature. Smaller spatial units are not uniformly available across the region and could be overly sensitive to limits in the spatial accuracy of many variables. Larger spatial units would decrease the ability to discern spatial patterns of environmental quality. Once the data set was assembled, a correlation analysis was performed to screen the data. The analysis identified variables that are highly related. High correlation coefficients may indicate that two variables are actually two different measurements of the same underlying ecological process. If this occurs, one of the variables can be dropped from the data set. In other cases, high correlation may simply indicate that high values of two variables tend to co-occur on the same watershed. If the variables measure different aspects of the underlying ecology, the variables are kept in the data set. However, it must be remembered that statistical analysis of the data set will be affected by these high correlations since they usually assume that the variables are independent. The next step was to code the data so that they could be combined in various ways. Some variables, such as number of exotic species, may range from 0 to 10. Other variables, such as human population density, may range from 50 to 2000 people per acre. To facilitate integration across different variables, ReVA codes all data from 0 to 1. A value of 0 indicates the best possible environmental conditions. Thus the highest value for each resource across the region is assigned a value of 0 and the lowest value of each stressor is assigned a value of 0. A value of 1 indicates the worst conditions. Thus, the lowest value for resources and the highest value for stressors are assigned a value of 1.0.
|
|
|
|
|
||
|
|