Location: Southeast Watershed ResearchTitle: Spatial definitions of working lands, LTAR croplands and grazinglands masks
|PONCE-CAMPOS, GUILLERMO - University Of Arizona|
|GAFFNEY, ROWAN - Unavco|
|Holifield Collins, Chandra|
|JAMES, DAVID - Iowa State University|
Submitted to: US-International Association for Landscape Ecology
Publication Type: Abstract Only
Publication Acceptance Date: 2/21/2022
Publication Date: 4/13/2022
Citation: Ponce-Campos, G., Coffin, A.W., Gaffney, R., Holifield Collins, C.D., James, D. 2022. Spatial definitions of working lands, LTAR croplands and grazinglands masks. US-International Association for Landscape Ecology. 4/13/2022; Virtual 2022.
Technical Abstract: The USDA-ARS Long-Term Agroecosystem Research (LTAR) Network is focused on developing national strategies for sustainable agriculture based on long-term research conducted across 18 sites that include croplands and grazinglands. To better understand the extent of these essential systems, we developed a spatial footprint, or “mask”, of these land use types in the conterminous United States(CONUS) for use as a key component in the data frameworks of the LTAR network. The methodology developed as part of this framework is based on the datasets available from the National Agricultural Statistics Service (NASS)-Crop Data Layer (CDL), for croplands, and National Land Cover Dataset (NLCD) for grazinglands. We used the available annual layers of each dataset to identify temporally stable pixels using the corresponding codes identified for land covers associated with each land use. For croplands, starting with 2016 as our base year, we classified all pixels as crop and non-crop. Then, we queried the other years to reclassify as croplands pixels those identified in 5 or more years as croplands. The result was a map of stable cropland pixels for a majority of years, including 2016. For grazinglands, a similar approach was followed using reclassified data from the NLCD (base year 2016), and 3 data years as part of the criteria to identify stable grazinglands pixels. A validation step, including local expert knowledge, was used to assess the classification accuracy of each layer and overall accuracy values were computed. The final layers were generated in geotiff format and made available for further analysis within the network.