Key Takeaways
- Kammerer et al. (2024) address the fragmentation of land cover data in the U.S. by integrating agricultural and natural vegetation datasets.
- The study merges LANDFIRE National Vegetation Classification (NVC) and USDA-NASS Cropland Data Layer (CDL) to create the ‘Spatial Products for Agriculture and Nature’ (SPAN).
- SPAN provides detailed rasters that encompass both agricultural and natural land-cover classes.
- The dataset, covering 2012-2021 for the conterminous United States, shows high accuracy, with only 0.6% unresolved agricultural pixels.
- With its comprehensive land cover characterization, SPAN is poised to benefit environmental research and management significantly.
Introduction to the Study On Land Cover Data
Bridging the Gap in Land Cover Data
Kammerer et al. (2024) undertake a crucial initiative to address a key challenge in environmental research: the fragmented nature of land cover data in the United States. Traditionally, datasets have focused either on agricultural or natural vegetation, but not both. This limitation has posed significant challenges for researchers and conservation practitioners who require a holistic landscape view for their studies and interventions.
Creation of SPAN
Merging Datasets for a Comprehensive View
To overcome this challenge, the researchers merged two significant datasets: the LANDFIRE National Vegetation Classification (NVC) and the USDA-NASS Cropland Data Layer (CDL). This integration led to the creation of ‘Spatial Products for Agriculture and Nature’ (SPAN), a tool that leverages the strengths of both datasets. SPAN offers detailed rasters that comprehensively cover agricultural and natural land-cover classes, a critical advancement for environmental studies.
Methodology and Timeframe
Detailed Rasters from 2012-2021
The SPAN dataset was generated annually over a decade, from 2012 to 2021, for the conterminous United States. This temporal range provides a valuable longitudinal perspective, allowing researchers to analyze trends and changes in land cover over time. The complete computational workflow used to create SPAN was also published, enhancing transparency and reproducibility.
Accuracy and Validation
High Agreement and Resolution of Conflicts
In their validation analyses, the researchers found that approximately 5.5% of NVC agricultural pixels conflicted with the CDL. However, most of these conflicts were effectively resolved, leaving only 0.6% of agricultural pixels unresolved in the final SPAN product. This high level of accuracy and agreement underscores the reliability of SPAN for environmental research and management purposes.
Implications for Environmental Research and Management
A Valuable Resource for Diverse Applications
With its integrated characterization of both agricultural and natural land cover, the SPAN dataset is expected to be a game-changer in environmental research and management. It provides a comprehensive landscape view, facilitating more informed and effective decision-making in land use planning, conservation strategies, and ecological studies.
Read more here.
Photo by Eric Brehm on Unsplash
