Environmental controls on seasonal ecosystem evapotranspiration/potential evapotranspiration ratio as determined by the global eddy flux measurements

Abstract

Abstract. The evapotranspiration / potential evapotranspiration (AET / PET) ratio is traditionally termed as the crop coefficient (K c ) and has been generally used as ecosystem evaporative stress index. In the current hydrology literature, K c has been widely used as a parameter to estimate crop water demand by water managers but has not been well examined for other types of ecosystems such as forests and other perennial vegetation. Understanding the seasonal dynamics of this variable for all ecosystems is important for projecting the ecohydrological responses to climate change and accurately quantifying water use at watershed to global scales. This study aimed at deriving monthly K c for multiple vegetation cover types and understanding its environmental controls by analyzing the accumulated global eddy flux (FLUXNET) data. We examined monthly K c data for seven vegetation covers, including open shrubland (OS), cropland (CRO), grassland (GRA), deciduous broad leaf forest (DBF), evergreen needle leaf forest (ENF), evergreen broad leaf forest (EBF), and mixed forest (MF), across 81 sites. We found that, except for evergreen forests (EBF and ENF), K c values had large seasonal variation across all land covers. The spatial variability of K c was well explained by latitude, suggesting site factors are a major control on K c . Seasonally, K c increased significantly with precipitation in the summer months, except in EBF. Moreover, leaf area index (LAI) significantly influenced monthly K c in all land covers, except in EBF. During the peak growing season, forests had the highest K c values, while croplands (CRO) had the lowest. We developed a series of multivariate linear monthly regression models for K c by land cover type and season using LAI, site latitude, and monthly precipitation as independent variables. The K c models are useful for understanding water stress in different ecosystems under climate change and variability as well as for estimating seasonal ET for large areas with mixed land covers.

Combining satellite data and appropriate objective functions for improved spatial pattern performance of a distributed hydrologic model

“…Regarding the dynamic scaling function, developed for incorporating remotely sensed LAI in ET ref scaling, it should be noted that the use of LAI to describe the deviation of each grid cell from the assumed reference grass is a simplification. Albedo could also have been included in the dynamic scaling function; however, one could argue that albedo and LAI are somewhat correlated and including one of them is already contributing the information about the other (Chen et al, 2005;Liu et al, 2017;Stisen et al, 2008). Moreover, we limit this 40 study to temporally averaged spatial patterns of AET and deliberately choose to ignore the day-to-day dynamics of AET.…”

“…The DSF has three dimensionless parameters that can be calibrated and a spatio-temporal LAI (-) component to account for the effects of characteristics that distinguish actual vegetation from reference grass (well-watered 10 cm height and having albedo of 0.23). These characteristics include specific land cover, albedo and aerodynamic resistance (Allen et al, 1998;Liu et al, 2017). This ensures a physically meaningful downscaling from a coarse (here 20 km) ET ref grid to the model resolution (here 1 km).…”

Combining satellite data and appropriate objective functions for improved spatial pattern performance of a distributed hydrologic model

“…Regarding the dynamic scaling function, developed for incorporating remotely sensed LAI in ET ref scaling, it should be noted that the use of LAI to describe the deviation of each grid cell from the assumed reference grass is a simplification. Albedo could also have been included in the dynamic scaling function; however, one could argue that albedo and LAI are somewhat correlated and including one of them is already contributing the information about the other (Chen et al, 2005;Liu et al, 2017;Stisen et al, 2008). Moreover, we limit this study to temporally averaged spatial patterns of AET and deliberately choose to ignore the day-to-day dynamics of AET.…”

“…(65), is simply a time-space variable implementation of the crop coefficient for natural vegetation, parameterized through a spatio-temporal LAI (no unit) component accounting for the effects of characteristics that separate the actual vegetation from a reference grass (well-watered 10 cm height and albedo of 0.23). These characteristics include specific land cover, albedo and aerodynamic resistance (Allen et al, 1998;Liu et al, 2017). This ensures a physically meaningful downscaling from a coarse (here 20 km) ET ref grid to the model resolution (here 1 km).…”

“…The multi-basin calibration results using MPR revealed physically meaningful parameter fields without patchiness (discontinuities). The study by Loosvelt et al (2013) is one of few other examples that incorporate PTFs for soil texture and moisture components of a hydrologic model.…”