Predictions In Basins Research Articles

Deep learning approaches for improving prediction of daily stream temperature in data‐scarce, unmonitored, and dammed basins

AbstractBasin‐centric long short‐term memory (LSTM) network models have recently been shown to be an exceptionally powerful tool for stream temperature (Ts) temporal prediction (training in one period and predicting in another period at the same sites). However, spatial extrapolation is a well‐known challenge to modelling Ts and it is uncertain how an LSTM‐based daily Ts model will perform in unmonitored or dammed basins. Here we compiled a new benchmark dataset consisting of >400 basins across the contiguous United States in different data availability groups (DAG, meaning the daily sampling frequency) with and without major dams, and studied how to assemble suitable training datasets for predictions in basins with or without temperature monitoring. For prediction in unmonitored basins (PUB), LSTM produced a root‐mean‐square error (RMSE) of 1.129°C and an R2 of 0.983. While these metrics declined from LSTM's temporal prediction performance, they far surpassed traditional models' PUB values, and were competitive with traditional models' temporal prediction on calibrated sites. Even for unmonitored basins with major reservoirs, we obtained a median RMSE of 1.202°C and an R2 of 0.984. For temporal prediction, the most suitable training set was the matching DAG that the basin could be grouped into (for example, the 60% DAG was most suitable for a basin with 61% data availability). However, for PUB, a training dataset including all basins with data was consistently preferred. An input‐selection ensemble moderately mitigated attribute overfitting. Our results indicate there are influential latent processes not sufficiently described by the inputs (e.g., geology, wetland covers), but temporal fluctuations can still be predicted well, and LSTM appears to be a highly accurate Ts modelling tool even for spatial extrapolation.

Maximizing the Potential of Artificial Intelligence to Perform Evaluations in Ungauged Washbowls

Long short-term memory networks (LSTM) offer precision in the prediction that has never been seen before in ungauged basins. Using k-fold validation, we trained and evaluated several LSTMs in this study on 531 basins from the CAMELS data set. This allowed us to make predictions in basins for which we did not have any training data. The implication is that there is usually sufficient information in available catchment attributes data about similarities and differences between catchment-level rainfall-runoff behaviors to generate out-of-sample simulations that are generally more accurate than current models when operating under ideal (i.e., calibrated) conditions, i.e., when using under idealized conditions. In other words, existing models are generally less accurate when working under idealized conditions than out-of-sample simulations. We found evidence that including physical limits in LSTM models improves simulations, which we believe should be the primary focus of future research on physics-guided artificial intelligence. Putting in place additional physical constraints on the LSTM models.

Application of fiducial method for streamflow prediction under small sample cases in Xiangxihe watershed, China

Hydrological prediction in basins with few data remains an important task for hydrologists given the practical relevance of hydrological prediction for water management and infrastructure design. In this article, fiducial method is applied for hydrological prediction to deal with small sample cases. Fiducial inference can be viewed as a procedure that obtains a measure on a parameter space while assuming less than Bayesian inference does (no prior); it can also be viewed, as a procedure that in a routine algorithmic way finds approximate pivots for parameters of interest, which is one of the main goals of frequential inference. In addition, fiducial methods require only the information from samples (streamflow data). Such that fiducial methods can account for valid information of streamflow observation and avoid model uncertainties. Three goodness-of-fit performance measures in terms of width of prediction interval (PI), accuracy and comprehensive measure will be examined to demonstrate the feasibility of fiducial method in hydrological prediction. Soil and water assessment tool (SWAT) based on Bayesian inference is used for comparison. Results show that (a) the performance of sharpness for the two methods is basically same under small nominal coverage; (b) fiducial PI can captures more observations with the similar width; (c) prediction performance of fiducial method is more satisfactory than SWAT based on Bayesian inference under small sample cases according to interval skill score; (d) fiducial method for hydrological prediction is much more time-saving than SWAT based on Bayesian inference. In summary, fiducial method has significant implications in increasing the performance and efficiency for hydrological prediction under small sample cases.

Enhancing Predictions in Ungauged Basins Using Machine Learning to Its Full Potential

In ungauged basins, long short-term memory (LSTM) networks provide unparalleled precision in prediction. Using k-fold validation, we trained and tested various LSTMs on 531 basins from the CAMELS data set, allowing us to make predictions in basins with no training data. The training and test data set contained 30 years of daily rainfall-runoff data from US catchments ranging in size from 4 to 2,000 km2, with aridity indexes ranging from 0.22 to 5.20, and 12 of the 13 IGPB vegetated land cover classes. Over a 15-year validation period, this effectively "ungauged" model was compared to the Sacramento Soil Moisture Accounting (SAC-SMA) model as well as the NOAA National Water Model reanalysis. Each basin's SAC-SMA was calibrated separately using 15 years of daily data. Across the 531 basins, the out-of-sample LSTM exhibited greater median Nash-Sutcliffe Efficiencies (0.69) than either the calibrated SAC-SMA (0.64) or the National Water Model (0.64). (0.58). This means that there is usually enough information in available catchment attributes data about similarities and differences between catchment-level rainfall-runoff behaviors to generate out-of-sample simulations that are generally more accurate than current models under ideal (i.e., calibrated) conditions. We discovered evidence that adding physical restrictions to the LSTM models improves simulations, which we believe should be the focus of future physics-guided machine learning research.