USDA National Agricultural Statistics Service Cropland Data Layer
The Cropland Data Layer (CDL) is a product of the USDA National Agricultural Statistics Service (NASS) with the mission “to provide timely, accurate and useful statistics in service to U.S. agriculture” (Johnson and Mueller, 2010, p. 1204). The CDL is a crop-specific land cover classification product of more than 100 crop categories grown in the United States. CDLs are derived using a supervised land cover classification of satellite imagery. The supervised classification relies on first manually identifying pixels within certain images, often called training sites, which represent the same crop or land cover type. Using these training sites, a spectral signature is developed for each crop type that is then used by the analysis software to identify all other pixels in the satellite image representing the same crop. Using this method, a new CDL is compiled annually and released to the public a few months after the end of the growing season through the online CropScape data portal.
The Cropland Data Layer (CDL) is a product of the USDA National Agricultural Statistics Service (NASS) with the mission “to provide timely, accurate and useful statistics in service to U.S. agriculture” (Johnson and Mueller, 2010, p. 1204). The CDL is a crop-specific land cover classification product of more than 100 crop categories grown in the United States. CDLs are derived using a supervised land cover classification of satellite imagery. The supervised classification relies on first manually identifying pixels within certain images, often called training sites, which represent the same crop or land cover type. Using these training sites, a spectral signature is developed for each crop type that is then used by the analysis software to identify all other pixels in the satellite image representing the same crop. Using this method, a new CDL is compiled annually and released to the public a few months after the end of the growing season through the online CropScape data portal (fig. 1).
The first CDL was produced in 1997 using data from Landsat 5 and covered three states: Arkansas, North Dakota, and South Dakota. Since then, the program has expanded to include coverage from additional satellites as well as the classification of additional states. By 2008, the annual CDL provided coverage for the entire continental United States and was being released at a 56m resolution. Additionally, several real-time CDL acreage estimates were being produced during the growing season for winter wheat, corn, and soybeans. Over the lifetime of the CDL, the program has thus far utilized a number of different earth observation satellites including Landsat 4/5/7, IRS-P6 Resourcesat-1 satellite, MODIS, DMC satellites Deimos-1 and UK-DMC 2, and as of 2013, Landsat 8.
Up until 2006, CDLs relied primarily on Landsat (5 and 7) data. An aging Landsat 5, the scan line corrector error (SLC-off) on Landsat 7, and uncertainty regarding the future launch of Landsat 8, forced the NASS to begin exploring alternative satellites to fill the pending data gap. The Indian Space Research Organization satellite Resourcesat-1, launched in 2003 carrying the AWiFS sensor, was selected as the most viable alternative. Bands for the AWiFS sensor were chosen to closely match those of the Landsat 7 Enhanced Thematic Mapper Plus (ETM+). AWiFS had a slightly lower resolution (56 meter) than Landsat Thematic Mapper (TM) and ETM+ (30 meter), but an increased revisit time of 5 days, which was beneficial given the dynamic nature of crops. Between 2007 and 2009, AWiFS provided the majority of imagery for CDLs. The 56-meter resolution of AWiFS, while adequate for mapping homogeneous crops such as soybean and corn, had a low accuracy for smaller, less homogenous crops. Beginning in 2010, CDLs were released at an updated 30m resolution and used a combination of Landsat TM/ETM+ and AWiFS (resampled to 30 meter using Landsat). The most recent CDL for 2013 was released January 2014 and relies primarily on Landsat 8 and DMC imagery.
The original development of the CDL was largely made possible through the growing availability of low cost and free mid-resolution Landsat-like imagery (Johnson and Mueller, 2010). The future of the CDL program is closely tied to the continuing availability of multi-spectral, mid-resolution imagery that is collected at a sufficient temporal frequency for crop forecasting. NASS utilizes CDLs as a principal component of crop yield modeling research and for developing real-time crop acreage estimates during the growing season. Monthly reports of area estimates are produced for June, August, twice in September, and October using Landsat and DMC data (Rick Mueller, USDA NASS, oral commun. and written commun., 2014). While the final CDL is not released until January the following year, monthly in-season estimates serve as the primary source for domestic production figures in the World Agricultural Supply and Demand Estimates (WASDE) report. Outside of NASS, CDLs have seen a growth in research and management applications including crop rotation, land use change, yield estimates, water use, and natural disaster impacts (Mueller and Harris, 2013; Boryan and Yang, 2013).
Benefits and Challenges of using Landsat Imagery
For NASS, using Landsat imagery to create CDLs provides an objective and unbiased assessment of farm-level crop production. The known stability of spectral and spatial resolution across Landsat satellites as well as the high level of image-data orthorectification and radiometric correction have helped establish Landsat as a gold standard among earth observation satellites. The thirty meter image-pixel resolution of the multi-spectral data collected by Landsat since 1982 has become a standard for land cover classification, enabling a high level of classification accuracy while balancing the processing realities of classifying imagery at a national scale (Johnson and Mueller, 2010). In practical terms, the availability of consistently collected (across multiple satellites/years), accurately corrected, and as of 2008, freely available, Landsat data is an important benefit for agencies which remain primarily operational in nature with limited time and budget for purchasing satellite imagery from commercial sources (Curt Reynolds, USDA FAS, written commun. and oral commun., 2014).
The recent launch of Landsat 8 in 2013 has helped ensure the continuity of the Landsat mission. However, between 2003 and 2013, the primary challenges facing agencies like the USDA were uncertainty in the long-term outlook of the Landsat program and the pending data gap due to an aging Landsat 5 and SLC-off issue with Landsat 7 (Johnson and Mueller, 2010). Although Landsat 8 became fully operational in 2013, other challenges related to agricultural monitoring remain, such as the 8-day repeat frequency, which is sometimes not often enough to obtain sufficient clear images in the growing season in certain areas.
References
Boryan, C., and Yang, Z., 2013, Deriving crop specific covariate data sets from multi-year NASS geospatial Cropland Data Layers in Geoscience and Remote Sensing Symposium (IGARSS), 2013 IEEE International: Melbourne, Australia, Institute of Electrical and Electronics Engineers, p. 4225–4228. doi:10.1109/IGARSS.2013.6723766.
Johnson, D., and Mueller, R., 2010, The 2009 Cropland Data Layer: Photogrammetric Engineering & Remote Sensing, v. 76, no. 11, p. 1201.
Mueller, R., and Harris, M., 2013, Reported uses of CropScape and the National Cropland Data Layer Program: Rio de Janeiro, Brazil, ICAS VI (Sixth International Conference on Agricultural Statistics), October 23-25, 2013, MuellerICASVI_CDL.pdf.
Case Studies of Landsat Imagery Use
The Cropland Data Layer (CDL) is a product of the USDA National Agricultural Statistics Service (NASS) with the mission “to provide timely, accurate and useful statistics in service to U.S. agriculture” (Johnson and Mueller, 2010, p. 1204). The CDL is a crop-specific land cover classification product of more than 100 crop categories grown in the United States. CDLs are derived using a supervised land cover classification of satellite imagery. The supervised classification relies on first manually identifying pixels within certain images, often called training sites, which represent the same crop or land cover type. Using these training sites, a spectral signature is developed for each crop type that is then used by the analysis software to identify all other pixels in the satellite image representing the same crop. Using this method, a new CDL is compiled annually and released to the public a few months after the end of the growing season through the online CropScape data portal.
The Cropland Data Layer (CDL) is a product of the USDA National Agricultural Statistics Service (NASS) with the mission “to provide timely, accurate and useful statistics in service to U.S. agriculture” (Johnson and Mueller, 2010, p. 1204). The CDL is a crop-specific land cover classification product of more than 100 crop categories grown in the United States. CDLs are derived using a supervised land cover classification of satellite imagery. The supervised classification relies on first manually identifying pixels within certain images, often called training sites, which represent the same crop or land cover type. Using these training sites, a spectral signature is developed for each crop type that is then used by the analysis software to identify all other pixels in the satellite image representing the same crop. Using this method, a new CDL is compiled annually and released to the public a few months after the end of the growing season through the online CropScape data portal (fig. 1).
The first CDL was produced in 1997 using data from Landsat 5 and covered three states: Arkansas, North Dakota, and South Dakota. Since then, the program has expanded to include coverage from additional satellites as well as the classification of additional states. By 2008, the annual CDL provided coverage for the entire continental United States and was being released at a 56m resolution. Additionally, several real-time CDL acreage estimates were being produced during the growing season for winter wheat, corn, and soybeans. Over the lifetime of the CDL, the program has thus far utilized a number of different earth observation satellites including Landsat 4/5/7, IRS-P6 Resourcesat-1 satellite, MODIS, DMC satellites Deimos-1 and UK-DMC 2, and as of 2013, Landsat 8.
Up until 2006, CDLs relied primarily on Landsat (5 and 7) data. An aging Landsat 5, the scan line corrector error (SLC-off) on Landsat 7, and uncertainty regarding the future launch of Landsat 8, forced the NASS to begin exploring alternative satellites to fill the pending data gap. The Indian Space Research Organization satellite Resourcesat-1, launched in 2003 carrying the AWiFS sensor, was selected as the most viable alternative. Bands for the AWiFS sensor were chosen to closely match those of the Landsat 7 Enhanced Thematic Mapper Plus (ETM+). AWiFS had a slightly lower resolution (56 meter) than Landsat Thematic Mapper (TM) and ETM+ (30 meter), but an increased revisit time of 5 days, which was beneficial given the dynamic nature of crops. Between 2007 and 2009, AWiFS provided the majority of imagery for CDLs. The 56-meter resolution of AWiFS, while adequate for mapping homogeneous crops such as soybean and corn, had a low accuracy for smaller, less homogenous crops. Beginning in 2010, CDLs were released at an updated 30m resolution and used a combination of Landsat TM/ETM+ and AWiFS (resampled to 30 meter using Landsat). The most recent CDL for 2013 was released January 2014 and relies primarily on Landsat 8 and DMC imagery.
The original development of the CDL was largely made possible through the growing availability of low cost and free mid-resolution Landsat-like imagery (Johnson and Mueller, 2010). The future of the CDL program is closely tied to the continuing availability of multi-spectral, mid-resolution imagery that is collected at a sufficient temporal frequency for crop forecasting. NASS utilizes CDLs as a principal component of crop yield modeling research and for developing real-time crop acreage estimates during the growing season. Monthly reports of area estimates are produced for June, August, twice in September, and October using Landsat and DMC data (Rick Mueller, USDA NASS, oral commun. and written commun., 2014). While the final CDL is not released until January the following year, monthly in-season estimates serve as the primary source for domestic production figures in the World Agricultural Supply and Demand Estimates (WASDE) report. Outside of NASS, CDLs have seen a growth in research and management applications including crop rotation, land use change, yield estimates, water use, and natural disaster impacts (Mueller and Harris, 2013; Boryan and Yang, 2013).
Benefits and Challenges of using Landsat Imagery
For NASS, using Landsat imagery to create CDLs provides an objective and unbiased assessment of farm-level crop production. The known stability of spectral and spatial resolution across Landsat satellites as well as the high level of image-data orthorectification and radiometric correction have helped establish Landsat as a gold standard among earth observation satellites. The thirty meter image-pixel resolution of the multi-spectral data collected by Landsat since 1982 has become a standard for land cover classification, enabling a high level of classification accuracy while balancing the processing realities of classifying imagery at a national scale (Johnson and Mueller, 2010). In practical terms, the availability of consistently collected (across multiple satellites/years), accurately corrected, and as of 2008, freely available, Landsat data is an important benefit for agencies which remain primarily operational in nature with limited time and budget for purchasing satellite imagery from commercial sources (Curt Reynolds, USDA FAS, written commun. and oral commun., 2014).
The recent launch of Landsat 8 in 2013 has helped ensure the continuity of the Landsat mission. However, between 2003 and 2013, the primary challenges facing agencies like the USDA were uncertainty in the long-term outlook of the Landsat program and the pending data gap due to an aging Landsat 5 and SLC-off issue with Landsat 7 (Johnson and Mueller, 2010). Although Landsat 8 became fully operational in 2013, other challenges related to agricultural monitoring remain, such as the 8-day repeat frequency, which is sometimes not often enough to obtain sufficient clear images in the growing season in certain areas.
References
Boryan, C., and Yang, Z., 2013, Deriving crop specific covariate data sets from multi-year NASS geospatial Cropland Data Layers in Geoscience and Remote Sensing Symposium (IGARSS), 2013 IEEE International: Melbourne, Australia, Institute of Electrical and Electronics Engineers, p. 4225–4228. doi:10.1109/IGARSS.2013.6723766.
Johnson, D., and Mueller, R., 2010, The 2009 Cropland Data Layer: Photogrammetric Engineering & Remote Sensing, v. 76, no. 11, p. 1201.
Mueller, R., and Harris, M., 2013, Reported uses of CropScape and the National Cropland Data Layer Program: Rio de Janeiro, Brazil, ICAS VI (Sixth International Conference on Agricultural Statistics), October 23-25, 2013, MuellerICASVI_CDL.pdf.