Species are categorized according to growth habit label of the USDA Plants Database.
Yes, both elevation and precipitation are used in estimating vegetation cover. Please see the dataset description.
Errors are described in the dataset description. These errors provide an accuracy assessment. In basic terms, vegetation cover estimates of a given pixel or region should be interpreted as plus or minus the error.
No, only vegetation cover is available.
The model uses seasonal summaries (e.g. spring maximums, summer means, etc.) of Landsat satellite data to estimate vegetation cover. In some years, limited satellite data are available over regions due to reduced satellite coverage, clouds, missing data, etc. This is particularly true for years 1984-1998 when only Landsat 5 was in orbit, but can occur in other years as well. These limited retrievals result in visual artifacts within the vegetation cover data.
Limited data in Landsat 7 retrievals also occur from year 2003 onward when the Scan Line Corrector (SLC) failed, resulting in data gaps throughout Landsat 7 scenes. These SLC gaps are often supplemented by subsequent passes of Landsat 7, data provided by Landsat 5 (in orbit until May 2012), and Landsat 8 (launched April 2013), but in some cases the effect of SLC gaps can be seen in the resulting vegetation cover estimates.
Spatial information and analysis results are not stored or archived. Only you can see the results of the analysis. All information is transported securely via industry standard encryption methods.
Downloadable maps and data are not available. Analysis results can be downloaded by clicking the save button next to the chart.
Yes. The Rangeland Analysis Platform is intended to be used alongside local knowledge and data. It is another tool for managers to consider when planning conservation and management actions.
Click the item within the legend to add/remove lines from the chart.
Yes. Please see shapefile help for more information.
Percent cover estimates are produced by combining over 30,000 field plots from the NRCS National Resources Inventory (NRI) and the BLM Assessment, Inventory, and Monitoring (AIM) datasets with the historical Landsat satellite record, gridded meteorology, and abiotic land surface data (e.g., elevation, soils). Utilizing the computation power of Google Earth Engine, cover is estimated at 30 m resolution, an area slightly larger than a baseball diamond.
Oftentimes, the region of interest (i.e., polygon) is not exactly aligned with what is happening on the ground. For example, uploading a shapefile of treatment polygons (tree removal, herbicide application, prescribed fire, etc.) may only represent planned treatment perimeters, but the actual treatment application may vary significantly within the planned perimeter. In this case, averaging across treated and untreated areas may yield a different estimate than averaging over only a treated area.
Additionally, many models that estimate cover have difficulty with low estimates and will rarely predict a cover of zero. It is important to inspect the area of analysis and to interpret the results in concert with local knowledge and data, and error metrics.