Satellite mapping of conservation tillage adoption in the Little River experimental watershed, Georgia

Abstract

Conservation tillage is a commonly adopted best management practice for improving soil quality and reducing erosion. However, there are currently no methods in place to monitor conservation tillage adoption at the watershed scale. The primary objective of this study was to evaluate the usefulness of Landsat TM data as a tool to depict conservation tillage in a small Coastal Plain watershed. Satellite imagery was used to calculate four commonly used indices: Normalized Difference Vegetation Index, Crop Residue Cover Index, Normalized Difference Tillage Index, and the Simple Tillage Index. Ground truth data consisted of a windshield survey, assigning each site a tillage regime (conventional or conservation tillage) at 138 locations throughout the watershed and surrounding areas. A logistical regression approach was used on two subsets of the data set (n = 20 or n = 44) to determine the influence of the number of ground control points on the success of modeling the occurrence of conservation tillage. The most accurate model was re-applied to the satellite image and evaluated using an independent sample of 94 survey sites. Results indicate that the normalized difference tillage and simple tillage indices performed best, with an overall accuracy of 71% and 78% for models developed using n = 20 and n = 44 sample locations, respectively. Errors were typically in the form of commission. Results are encouraging and suggest that currently available satellite imagery can be used for rapid assessment of conservation tillage adoption using minimal a priori information.

Mapping Crop Residue by Combining Landsat and WorldView-3 Satellite Imagery

“…The spectral profiles of the Landsat bands are as follows: Blue (0.45-0.52 µm), Green (0.52-0.60 µm), Red (0.63-0.69 µm), NIR (0.77-0.90 µm), SWIR1 (1.55-1.75 µm), and SWIR2 (2.09-2.35 µm), corresponding respectively to Landsat 8 (OLI) bands 2-7 or Landsat 7 (ETM+) bands 1-5, and 7. The indices included the Normalized Difference Indices 5 and 7 (NDI5 [29], NDI7 [29]), the Normalized Difference Senescent Vegetation Index (NDSVI [30]), Normalized Difference Tillage Index (NDTI [13]), the Simple Tillage Index (STI [13]), and the Modified Crop Residue Cover (mCRC) index [1,12]. (SWIR1 − Green)/(SWIR1 + Green) Sullivan et al [1,12] Four green vegetation indices were also tested for inclusion in the analysis, including the Normalized Difference Vegetation Index (NDVI) [31], Optimized Soil-Adjusted Vegetation Index (OSAVI) [32], Visible Atmospherically Resistant Index (VARI) [33], and the Green Vegetation Index (VIGreen) [34,35].…”

Multitemporal remote sensing of crop residue cover and tillage practices: A validation of the minNDTI strategy in the United States

“…Of these, the Normalized Difference Tillage Index (NDTI) (van Deventer et al 1997) was found to be the most effective (Serbin et al 2009b(Serbin et al , 2009a), but was not as effective as CAI, SINDRI, or LCA. In contrast, others concluded that these tillage indices were able to differentiate conventional and conservation tillage using logistic regression (Gowda et al 2001;Sullivan et al 2008;van Deventer et al 1997). Zheng et al (2012) found that these mixed results were caused by inattention to the significance of the temporal dimensions of applications of tillage practices.…”

Performance of the Soil Vulnerability Index with respect to slope, digital elevation model resolution, and hydrologic soil group

“…Most slopes were less than 5%, with some valley side slopes ranging from 5% to 15% (Bosch et al 2007). Land use in the watershed included agriculture (31%), pasture (10%), riparian forest (28%), upland forest (22%), and urban area (7%) (Sullivan et al 2008). The primary crops grown were row crops (cotton, corn, and peanut [Arachis hypogea L.]), and vegetables, typically without artificial drainage.…”