Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/1/1997
Publication Date: N/A
Citation: N/A Interpretive Summary: Knowledge of soil moisture is useful for important farm management practices such as yield prediction and irrigation scheduling. Conventional methods for measuring soil moisture data are limited by intensive labor costs. As an alternative, airborne radiometers could provide frequent estimates of soil moisture over large areas at low cost. However, the accuracy of such indirect soil moisture estimates is influenced by soil surface roughness conditions and the amount of vegetation covering the soil. In this study, the sensitivity of remotely-sensed data to soil moisture was evaluated over agricultural fields characterized by level-basin or furrow irrigation systems. The results showed that remote sensing could be used to estimate soil moisture, but the accuracy was dependent upon the sensor characteristics as well as field conditions. Taking these factors into account, this remote approach for estimating soil lmoisture can be applied to other agricultural fields with similar soil tillage practices. Such information will be used by farm managers to improve crop yield and decrease farm costs.
Technical Abstract: A sensitivity analysis of C-band (5.3 GHz) and Ku-bank (14.85 GHz) synthetic aperture radar (SAR) data to the bare soil moisture content of agricultural fields was conducted in this study. The C-band data were obtained with a 23-degree incidence angle, whereas the Ku-band data were obtained with 35-degree, 55-degree, and 75- degree incidence angles. The fields presented either a small-scale or an intermediate-scale periodic soil roughness component, associated with level-basin and furrow irrigation systems, respectively. For fields with a small-scale roughness component, the SAR data were sensitive to the soil moisture, particularly at the C-band with a 23-degree incidence angle and Ku-band with a 35-degree incidence angle. For fields with an intermediate-scale roughness component, both C- and Ku-band data were nearly insensitive to the soil moisture. By using a theoretical surface scattering model, this study also analyzed the effects of different soil roughness components [root mean square (RMS) height h, correlation length , and periodic row structure] in the SAR data. For fields with RMS height <0.3 cm, a small variation in h (from 0.1 to 0.3 cm) provoked a significant variation in the SAR data (up to 8 dB).