This study developed and evaluated 30‐m daily evapotranspiration (ET) estimates using the Priestley‐Taylor Jet Propulsion Laboratory (PT‐JPL) model with ECOSTRESS, Moderate MODIS, harmonized Landsat Sentinel‐2 (HLS) imagery, ERA5‐Land reanalysis, and eddy covariance measurements. The new daily 30‐m ET showed significantly improved performance (overall,
Characterization of model errors is important when applying satellite-driven evapotranspiration (ET) models to water resource management problems. This study examines how uncertainty in meteorological forcing data and land surface modeling propagate through to errors in final ET data calculated using the Satellite Irrigation Management Support (SIMS) model, a computationally efficient ET model driven with satellite surface reflectance values. The model is applied to three instrumented winegrape vineyards over the 2017–2020 time period and the spatial and temporal variation in errors are analyzed. We illustrate how meteorological data inputs can introduce biases that vary in space and at seasonal timescales, but that can persist from year to year. We also observe that errors in SIMS estimates of land surface conductance can have a particularly strong dependence on time of year. Overall, meteorological inputs introduced RMSE of 0.33–0.65 mm/day (7–27%) across sites, while SIMS introduced RMSE of 0.55–0.83 mm/day (19–24%). The relative error contribution from meteorological inputs versus SIMS varied across sites; errors from SIMS were larger at one site, errors from meteorological inputs were larger at a second site, and the error contributions were of equal magnitude at the third site. The similar magnitude of error contributions is significant given that many satellite-driven ET models differ in their approaches to estimating land surface conductance, but often rely on similar or identical meteorological forcing data. The finding is particularly notable given that SIMS makes assumptions about the land surface (no soil evaporation or plant water stress) that do not always hold in practice. The results of this study show that improving SIMS by eliminating these assumptions would result in meteorological inputs dominating the error budget of the model on the whole. This finding underscores the need for further work on characterizing spatial uncertainty in the meteorological forcing of ET.
more » « less- NSF-PAR ID:
- 10372053
- Publisher / Repository:
- Springer Science + Business Media
- Date Published:
- Journal Name:
- Irrigation Science
- Volume:
- 40
- Issue:
- 4-5
- ISSN:
- 0342-7188
- Page Range / eLocation ID:
- p. 515-530
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract r = 0.8, RMSE = 1.736, KGE = 0.466) at 145 EC sites over contiguous United States compared to the current 70‐m ECOSTRESS ET (overall,r = 0.485, RMSE = 4.696, KGE = −0.841). A deep neural network postprocessing model trained with ET measurements from EC sites further improved the performance on test sites that were not used for model training (overall,r = 0.842, RMSE = 0.88, KGE = 0.792). The 30‐m ET estimation biases were significantly related to the biases in the upwelling longwave (R UL ) and downwelling shortwave radiation (R DS ) inputs, with ET estimates driven by MODIS radiation showing higher biases compared to those driven by ERA5‐Land radiation. The error diagnosis using random forest indicates that ET biases tend to be larger under higher ET estimates, andR UL andR DS were the primary contributors to the high bias at the higher ET ranges, with partial dependence plots revealing that the estimation biases tend to be higher under more humid environment, denser vegetation covers, and high net radiation conditions. In conclusion, higher spatial resolution satellite imagery of vegetation characteristics and higher temporal resolution radiation data, combined with continent‐wide EC measurements and deep learning, provided substantial added value for improving ET estimations at the field scale (30‐m). -
more » « less
We propose a conceptual and theoretical foundation for information-based model benchmarking and process diagnostics that provides diagnostic insight into model performance and model realism. We benchmark against a bounded estimate of the information contained in model inputs to obtain a bounded estimate of information lost due to model error, and we perform process-level diagnostics by taking differences between modeled versus observed transfer entropy networks. We use this methodology to reanalyze the recent Protocol for the Analysis of Land Surface Models (PALS) Land Surface Model Benchmarking Evaluation Project (PLUMBER) land model intercomparison project that includes the following models: CABLE, CH-TESSEL, COLA-SSiB, ISBA-SURFEX, JULES, Mosaic, Noah, and ORCHIDEE. We report that these models (i) use only roughly half of the information available from meteorological inputs about observed surface energy fluxes, (ii) do not use all information from meteorological inputs about long-term Budyko-type water balances, (iii) do not capture spatial heterogeneities in surface processes, and (iv) all suffer from similar patterns of process-level structural error. Because the PLUMBER intercomparison project did not report model parameter values, it is impossible to know whether process-level error patterns are due to model structural error or parameter error, although our proposed information-theoretic methodology could distinguish between these two issues if parameter values were reported. We conclude that there is room for significant improvement to the current generation of land models and their parameters. We also suggest two simple guidelines to make future community-wide model evaluation and intercomparison experiments more informative.
-
more » « less
Forecasting the timing and magnitude of snowmelt and runoff is critical to managing mountain water resources. Warming temperatures are increasing the rain–snow transition elevation and are limiting the forecasting skill of statistical models relating historical snow water equivalent to streamflow. While physically based methods are available, they require accurate estimations of the spatial and temporal distribution of meteorological variables in complex terrain. Across many mountainous areas, measurements of precipitation and other meteorological variables are limited to a few reference stations and are not adequate to resolve the complex interactions between topography and atmospheric flow. In this paper, we evaluate the ability of the Weather Research and Forecasting (WRF) Model to approximate the inputs required for a physics-based snow model, iSnobal, instead of using meteorological measurements, for the Boise River Basin (BRB) in Idaho, United States. An iSnobal simulation using station data from 40 locations in and around the BRB resulted in an average root-mean-square error (RMSE) of 4.5 mm compared with 12 SNOTEL measurements. Applying WRF forcings alone was associated with an RMSE of 10.5 mm, while including a simple bias correction to the WRF outputs of temperature and precipitation reduced the RMSE to 6.5 mm. The results highlight the utility of using WRF outputs as input to snowmelt models, as all required input variables are spatiotemporally complete. This will have important benefits in areas with sparse measurement networks and will aid snowmelt and runoff forecasting in mountainous basins.
-
more » « less
Abstract Accurate prediction of snow water equivalent (SWE) can be valuable for water resource managers. Recently, deep learning methods such as long short-term memory (LSTM) have exhibited high accuracy in simulating hydrologic variables and can integrate lagged observations to improve prediction, but their benefits were not clear for SWE simulations. Here we tested an LSTM network with data integration (DI) for SWE in the western United States to integrate 30-day-lagged or 7-day-lagged observations of either SWE or satellite-observed snow cover fraction (SCF) to improve future predictions. SCF proved beneficial only for shallow-snow sites during snowmelt, while lagged SWE integration significantly improved prediction accuracy for both shallow- and deep-snow sites. The median Nash–Sutcliffe model efficiency coefficient (NSE) in temporal testing improved from 0.92 to 0.97 with 30-day-lagged SWE integration, and root-mean-square error (RMSE) and the difference between estimated and observed peak SWE values
d maxwere reduced by 41% and 57%, respectively. DI effectively mitigated accumulated model and forcing errors that would otherwise be persistent. Moreover, by applying DI to different observations (30-day-lagged, 7-day-lagged), we revealed the spatial distribution of errors with different persistent lengths. For example, integrating 30-day-lagged SWE was ineffective for ephemeral snow sites in the southwestern United States, but significantly reduced monthly-scale biases for regions with stable seasonal snowpack such as high-elevation sites in California. These biases are likely attributable to large interannual variability in snowfall or site-specific snow redistribution patterns that can accumulate to impactful levels over time for nonephemeral sites. These results set up benchmark levels and provide guidance for future model improvement strategies. -
more » « less
Abstract Surface‐atmosphere fluxes and their drivers vary across space and time. A growing area of interest is in downscaling, localizing, and/or resolving sub‐grid scale energy, water, and carbon fluxes and drivers. Existing downscaling methods require inputs of land surface properties at relatively high spatial (e.g., sub‐kilometer) and temporal (e.g., hourly) resolutions, but many observed land surface drivers are not continuously available at these resolutions. We evaluate an approach to overcome this challenge for land surface temperature (LST), a World Meteorological Organization Essential Climate Variable and a key driver for surface heat fluxes. The Chequamegon Heterogenous Ecosystem Energy‐balance Study Enabled by a High‐density Extensive Array of Detectors (CHEESEHEAD19) field experiment provided a scalable testbed. We downscaled LST from satellites (GOES‐16 and ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station [ECOSTRESS]) with further refinement using airborne hyperspectral imagery. Temporally and spatially downscaled LST compared well to independent observations from a network of 20 micrometeorological towers and piloted aircrafts in addition to Landsat‐based LST retrieval and drone‐based LST observed at one tower site. The downscaled 50‐m hourly LST showed good relationships with tower (
r 2 = 0.79, RMSE = 3.5 K) and airborne (r 2 = 0.75, RMSE = 2.4 K) observations over space and time, with precision lower over wetlands and lakes, and some improvement for capturing spatio‐temporal variation compared to a geostationary satellite. Further downscaling to 10 m using hyperspectral imagery resolved hot and cold spots across the landscape as evidenced by independent drone LST, with significant reduction in RMSE by 1.3 K. These results demonstrate a simple pathway for multi‐sensor retrieval of high space and time resolution LST.
