Integration of an unmanned aircraft system and ground-based remote sensing to estimate spatially distributed crop evapotranspiration and soil water deficit throughout the vegetation soil root zone.

摘要

Irrigation is the largest consumer of fresh water and produces over 40% of the world's food and fiber supply. As the world's population continues to grow rapidly, the increased demands on fresh water will force the agricultural community to improve the efficiency and productivity of irrigation systems, while reducing overall water usage. In order to address the requirements of increased efficiency and productivity in agricultural water use, the agricultural community has begun to focus on the development of precision agriculture (PA) irrigation management systems for use with irrigated agriculture. Remote sensing (RS) is at the forefront of the PA movement, allowing the estimation of spatially distributed crop water requirements on a large-scale basis. Techniques using ground, aerial and space-borne RS platforms, have been developed to estimate actual crop evapotranspiration (ETa) and soil water deficit (SWD) for use in PA irrigation management systems. The ability to monitor the ET a and SWD allows irrigators to manage their irrigation to increase efficiency and decrease overall water use while maintaining crop yields goals. Historically, remote sensing data, such as spectral reflectance and thermal infrared (TIR) imagery, were provided by ground or space-borne RS platforms, like NASA's Landsat 8 satellites. Though these methods are effective at estimating ET a over large areas, their lack of spatial and temporal resolution limit their effectiveness for application in PA irrigation management systems. In order to address the required spatial and temporal resolutions required for PA systems, Colorado State University (CSU) developed an unmanned aircraft system (UAS) RS platform capable of collecting high spatial and temporal resolution data in the TIR, near-infrared (NIR), red and green bands of the electromagnetic spectrum. During the summer of 2015, CSU conducted four flights over corn at the Agriculture Research Development and Education Center (ARDEC), near Fort Collins, CO, with the Tempest UAS RS platform in order to collect thermal and multispectral imagery. The RS data collected over the ARDEC test location were used in three studies. The first was the comparison of the raw RS data to the ground-based RS data collected during the RS overpasses. The second study used the Tempest RS data to estimate the ETa using four methods: two methods based on the surface energy balance (Two-Source Energy Balance (TSEB) and the Surface Aerodynamic Temperature (SAT)), one method based on the TIR imagery (Crop Water Stress Index (CWSI)), and one method based on the spectral reflectance imagery (reflectance-based crop coefficients (k cbrf)) and reference ET. Remote sensing derived ET a estimates were compared to ETa derived using neutron probe soil moisture sensors. The third study utilized the RS derived ETa and the Hybrid Soil Water Balance method to estimate the SWD for comparison with the neutron probe derived SWD. Results showed that the Tempest RS data was in good agreement with the ground-based data as demonstrate by the low RMSE of the raw data, ETa and SWD calculations (TIR = 5.68 °C, NIR = 5.26 % reflectance, red = 3.51 % reflectance, green = 7.31 % reflectance, TSEB ETa = 0.89 mm/d, Hybrid SWD = 16.19 mm/m). The accuracy of the results of the Tempest UAS RS platform suggests that UAS RS platforms have the potential to increase the accuracy of ETa¬ and SWD estimation for use in the application of a PA irrigation management system.