| Remote sensing (RS) may be used to rapidly assess surface features and facilitate natural resource management, precision agriculture and soil survey. Information obtained in such a way would streamline data collection and improve diagnostic capabilities. Current RS technology has had limited testing, particularly within the Southeast. Our study was designed to evaluate RS as a rapid assessment tool in three different natural resource applications: nitrogen (N) management in a corn crop (Zea mays L.), assessment of in situ crop residue cover, and quantification of near-surface soil properties.; In 2000, study sites were established in four physiographic provinces of Alabama: Tennessee Valley, Ridge and Valley, Appalachian Plateau, and Coastal Plain. Spectral measurements were acquired via spectroradiometer (350–1050 nm), airborne ATLAS multispectral scanner (400–12,500 nm), and IKONOS satellite (450–900 nm).; Corn plots were established from fresh-tilled ground in a completely randomized design at the Appalachian Plateau and Coastal Plain study sites in 2000. Plots received four N rates (0, 56, 112, and 168 kg N ha−1 ), and were maintained for three consecutive growing seasons. Spectroradiometer data were acquired biweekly from V6-R2 and ATLAS and IKONOS were acquired per availability. Results showed vegetation indices derived from hand-held spectroradiometer measurements as early as V6-V8 were linearly related to yield and tissue N. ATLAS imagery showed promise at the AP site during the V6 stage (r2 = 0.66), but no significant relationships between plant N and IKONOS imagery were observed.; Residue plots (15m x 15m) were established at the Appalachian Plateau and Coastal Plain in 2000 and 200. Residue treatments consisted of hand applied wheat straw cover (0, 10 20, 50, or 80%) arranged in a completely randomized design. Spectroradiometer data were acquired monthly and ATLAS and IKONOS were acquired per availability. Residue cover estimates were best with ATLAS data explaining up to 98% of the variability, followed by spectroradiometer 84%, color infrared photography 56%, and IKONOS 24%.; In the third study, surface soil samples were collected at random from each of the four physiographies (163 samples) and analyzed for soil texture, citrate-dithionite extractable iron (Fed), soil organic carbon (SOC), soil water content, roughness and crusting. Remotely sensed data were acquired via the Airborne Terrestrial Applications Sensor in 2000. Results showed RS data acquired from arable lands with less than 4% surface soil water content, under our study constraints, best approximated near-surface soil properties. Results were best at the Coastal Plain site where loamy sand textured surfaces were predominant. Utilizing a combination of band ratios in stepwise regression Fed (r2 = 0.61), SOC (r2 = 0.36), sand (r2 = 0.52), and clay (r2 = 0.76) were related to RS data at the Coastal Plain site. Emissivity estimates did not generally improve estimates of near-surface soil attributes. |
