Multispectral remote sensing to estimate actual crop coefficients and evapotranspiration rates for grass pastures in western Colorado
Date
2018
Authors
Gautam, Sumit, author
Chávez, José L., advisor
Cabot, Perry, advisor
Brummer, Joe, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Evapotranspiration is the process by which water moves into the atmosphere by evaporation from the soil surface and transpiration from growing plants. Knowledge of crop evapotranspiration (ETc) is important for effective irrigation water management. Among the various methods used to estimate ETc, the standardized FAO56 Penman-Monteith approach, using tabulated generalized Kc values, has been widely adopted to estimate crop evapotranspiration. Remote sensing techniques are growing rapidly as a way to monitor actual crop water use. Remotely sensed data are used in algorithms to measure the spectral reflectance of the crop canopies. The differences in reflectance values, at different bandwidths from typical multispectral signatures, help determine the current or actual canopy properties like fractional crop cover, water stress, nutrient level, etc. The actual crop coefficients (Kca) were calculated using actual crop evapotranspiration (ETa) and alfalfa based reference crop evapotranspiration (ETr) rates. The soil water balance approach was used to calculate ETa for grass hay/pasture during the 2016 and 2017 growing seasons. A handheld multispectral radiometer was used to collect surface/canopy reflectance data. Vegetation indices (VI) were calculated using the surface reflectance data. Vegetation indices are the mathematical combination or transformation of surface reflectance in different spectral bands. Vegetation indices were then related to Kca to develop Kca(VI) models. Among the 11 different Kca(VI) based models evaluated, the Green normalized difference vegetation index (GNDVI) based Kca(VI) model performed better on a daily timestep. Depending upon the availability of surface reflectance readings, the user can use either of the four Kca(VI) based models: GNDVI, Transformed vegetation index (TVI), Normalized difference vegetation index (NDVI), or Infrared percentage vegetation index (IPVI) to estimate ETa. However, it is recommended to use the GNDVI based Kca(VI) model for increased accuracy. The results from this study can be used to estimate near real-time ETa rates for grass hay/pastures.