Irrigation Water Management Using Remote Sensing and Hydrologic Modeling
Abstract
Irrigation is the primary water source for agricultural production in arid and semi-arid regions for reliable food production and reducing risk due to irregular rainfall. Currently, few growers are scheduling their irrigation systems based on soil moisture sensor or weather station observations, rather most growers schedule irrigation based on non-scientific methods including visual or the soil feel method. Hydrologic models and remote sensing may have a significant impact on irrigation water management practices, since they summarize the entire hydrologic cycle by considering water sources (precipitation, surface, subsurface), amount of water added, evapotranspiration, and crop yield. Two study sites were selected to conduct this study, the ALG watershed in north east Indiana with humid conditions and the USDA-ARS Conservation and Production Research Laboratory (CPRL) at Bushland, Texas with semi-arid conditions. The overall objective of this research is to determine whether it is possible to fully depend on remote sensing in irrigation water management, or whether hydrological models should be considered as well for more accurate estimation of crop water requirements and soil moisture estimation. The satellite and hydrologic model evapotranspiration estimates were evaluated for both sites, and the hydrologic model soil moisture estimates were evaluated using daily and weekly soil moisture observation for the Indiana and Texas sites, respectively. Based on the results of this research, hydrologic model ET estimates under dryland conditions are not acceptable with the default leaf area index (LAI) but were acceptable with the modified LAI and for irrigated conditions. Landsat was able to estimate ET and LAI accurately for irrigated fields, and large uncertainties occur with the dryland fields due to water stress that reduces LAI. MODIS tends to under predict the ET compared to Landsat due to bigger pixel size, different estimation technique and different potential ET equation used. The hydrologic model soil moisture estimates show large errors for both sites due to model uncertainties and inappropriate soil data parameters. Based on study findings, Landsat and hydrologic modeling ET estimates can be used for irrigation management and agricultural water use estimation to some extent for irrigated fields, however limitations apply for dryland fields. Hydrologic model soil moisture estimates would be considered unacceptable for real time irrigation management purposes.
Degree
Ph.D.
Advisors
Engel, Purdue University.
Subject Area
Agricultural engineering|Engineering|Agriculture|Hydrologic sciences|Water Resources Management
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.