Opportunities and challenges for the use of scintillometer-based catchment-averaged evapotranspiration estimates as model forcing

Abstract. To date, lumped rainfall-runoff models rely on rough estimates of catchment-averaged potential evapotranspiration (ETp) rates as meteorological forcing. A model parameter converts this ETp input into actual evapotranspiration (ETact) estimates. This paper examines the potential use of scintillometer-based ETact rates for rainfall-runoff modeling. It has been found that the reservoir-structure of the rainfall-runoff model functions as a low-pass filter for the ETp input. If the long-term volume of the ETp used in the model simulations is consistent with the data set used for calibration, a good match of the seasonal pattern, using temporally constant ETp data, is sufficient to obtain adequate discharge simulations. However, these results are then obtained with strongly erroneous evapotranspiration estimates. A better match of the diurnal cycle does not lead to better model results. Replacing the ETp inputs by scintillometer-based ETact estimates does not lead to better model predictions. Small underestimations of ETact under stable conditions, which occur at night and during the Winter, and which accumulate to significant amounts, are the cause of this problem. Consistent with other studies, the scintillometer-based ETact estimates can be considered reliable and realistic under unstable conditions. These values can thus be used as forcing for rainfall-runoff models.

Download Full-text

A note on leveraging synergy in multiple meteorological datasets with deep learning for rainfall-runoff modeling

10.5194/hess-2020-221 ◽

2020 ◽

Author(s):

Frederik Kratzert ◽

Daniel Klotz ◽

Sepp Hochreiter ◽

Grey S. Nearing

Keyword(s):

Deep Learning ◽

Short Term Memory ◽

Rainfall Runoff ◽

Data Set ◽

Simulation Accuracy ◽

Meteorological Forcing ◽

Runoff Modeling ◽

Long Short Term Memory ◽

Rainfall Runoff Model ◽

Runoff Model

Abstract. A deep learning rainfall-runoff model can take multiple meteorological forcing products as inputs and learn to combine them in spatially and temporally dynamic ways. This is demonstrated using Long Short Term Memory networks (LSTMs) trained over basins in the continental US using the CAMELS data set. Using multiple precipitation products (NLDAS, Maurer, DayMet) in a single LSTM significantly improved simulation accuracy relative to using only individual precipitation products. A sensitivity analysis showed that the LSTM learned to utilize different precipitation products in different ways in different basins and for simulating different parts of the hydrograph in individual basins.

Download Full-text

Impact of potential and (scintillometer-based) actual evapotranspiration estimates on the performance of a lumped rainfall–runoff model

Hydrology and Earth System Sciences ◽

10.5194/hess-17-4525-2013 ◽

2013 ◽

Vol 17 (11) ◽

pp. 4525-4540 ◽

Cited By ~ 10

Author(s):

B. Samain ◽

V. R. N. Pauwels

Keyword(s):

Water Balance ◽

Stream Flow ◽

Seasonal Pattern ◽

Early Spring ◽

Actual Evapotranspiration ◽

Rainfall Runoff ◽

Forecasting Models ◽

Stable Conditions ◽

Rainfall Runoff Model ◽

Runoff Model

Abstract. Evapotranspiration (ET) plays a key role in hydrological impact studies and operational flood forecasting models as ET represents a loss of water from a catchment. Although ET is a major component of the catchment water balance, the evapotranspiration input for rainfall–runoff models is often simplified in contrast to the detailed estimates of catchment averaged precipitation. In this study, an existing conceptual rainfall–runoff model calibrated for and operational in the Bellebeek catchment in Belgium firstly has been validated and its sensitivity to different available potential ET input has been studied. It has been shown that when applying a calibrated rainfall–runoff model, the model input should be consistent with the input used for the calibration process, not only on the volume of ET, but also on the seasonal pattern. Secondly, estimates of the actual evapotranspiration based on measurements of a large aperture scintillometer (LAS) have been used as model forcing in the rainfall–runoff model. From this analysis, it has been shown that the actual evapotranspiration is a crucial factor in simulating the catchment water balance and the resulting stream flow. Regarding the actual evapotranspiration estimates from the LAS, it has been concluded that they can be considered realistic in summer months. In the months where stable conditions prevail (autumn, winter and (early) spring), an underestimation of the actual evapotranspiration is made, which has an important impact on the catchment's water balance.

Download Full-text

A note on leveraging synergy in multiple meteorological data sets with deep learning for rainfall–runoff modeling

Hydrology and Earth System Sciences ◽

10.5194/hess-25-2685-2021 ◽

2021 ◽

Vol 25 (5) ◽

pp. 2685-2703

Author(s):

Frederik Kratzert ◽

Daniel Klotz ◽

Sepp Hochreiter ◽

Grey S. Nearing

Keyword(s):

Deep Learning ◽

Short Term Memory ◽

Meteorological Data ◽

Data Sets ◽

Rainfall Runoff ◽

Data Set ◽

Simulation Accuracy ◽

Meteorological Forcing ◽

Rainfall Runoff Model ◽

Land Data Assimilation System

Abstract. A deep learning rainfall–runoff model can take multiple meteorological forcing products as input and learn to combine them in spatially and temporally dynamic ways. This is demonstrated with Long Short-Term Memory networks (LSTMs) trained over basins in the continental US, using the Catchment Attributes and Meteorological data set for Large Sample Studies (CAMELS). Using meteorological input from different data products (North American Land Data Assimilation System, NLDAS, Maurer, and Daymet) in a single LSTM significantly improved simulation accuracy relative to using only individual meteorological products. A sensitivity analysis showed that the LSTM combines precipitation products in different ways, depending on location, and also in different ways for the simulation of different parts of the hydrograph.

Download Full-text

Assessing the impact of PET estimation methods on hydrologic model performance

Hydrology Research ◽

10.2166/nh.2020.066 ◽

2020 ◽

Author(s):

Dilhani Ishanka Jayathilake ◽

Tyler Smith

Keyword(s):

Potential Evapotranspiration ◽

Model Performance ◽

Monte Carlo Sampling ◽

Hydrologic Model ◽

Estimation Methods ◽

Rainfall Runoff ◽

Data Set ◽

Rainfall Runoff Model ◽

The Impact ◽

Runoff Model

Abstract Evapotranspiration is a necessary input and one of the most uncertain hydrologic variables for quantifying the water balance. Key to accurately predicting hydrologic processes, particularly under data scarcity, is the development of an understanding of the regional variation of the impact of potential evapotranspiration (PET) data inputs on model performance and parametrization. This study explores this impact using four different potential evapotranspiration products (of varying quality). For each data product, a lumped conceptual rainfall–runoff model (GR4J) is tested on a sample of 57 catchments included in the MOPEX data set. Monte Carlo sampling is performed, and the resulting parameter sets are analyzed to understand how the model responds to differences in the forcings. Test catchments are classified as energy- or water-limited using the Budyko framework and by eco-region, and the results are further analyzed. While model performance (and parameterization) in water-limited sites was found to be largely unaffected by the differences in the evapotranspiration inputs, in energy-limited sites model performance was impacted as model parameterizations were clearly sensitive to evapotranspiration inputs. The quality/reliability of PET data required to avoid negatively impacting rainfall–runoff model performance was found to vary primarily based on the water and energy availability of catchments.

Download Full-text