Evaluating post-processing approaches for monthly and seasonal streamflow forecasts

Abstract

Abstract. Streamflow forecasting is prone to substantial uncertainty due to errors in meteorological forecasts, hydrological model structure, and parameterization, as well as in the observed rainfall and streamflow data used to calibrate the models. Statistical streamflow post-processing is an important technique available to improve the probabilistic properties of the forecasts. This study evaluates post-processing approaches based on three transformations – logarithmic (Log), log-sinh (Log-Sinh), and Box–Cox with λ=0.2 (BC0.2) – and identifies the best-performing scheme for post-processing monthly and seasonal (3-months-ahead) streamflow forecasts, such as those produced by the Australian Bureau of Meteorology. Using the Bureau's operational dynamic streamflow forecasting system, we carry out comprehensive analysis of the three post-processing schemes across 300 Australian catchments with a wide range of hydro-climatic conditions. Forecast verification is assessed using reliability and sharpness metrics, as well as the Continuous Ranked Probability Skill Score (CRPSS). Results show that the uncorrected forecasts (i.e. without post-processing) are unreliable at half of the catchments. Post-processing of forecasts substantially improves reliability, with more than 90 % of forecasts classified as reliable. In terms of sharpness, the BC0.2 scheme substantially outperforms the Log and Log-Sinh schemes. Overall, the BC0.2 scheme achieves reliable and sharper-than-climatology forecasts at a larger number of catchments than the Log and Log-Sinh schemes. The improvements in forecast reliability and sharpness achieved using the BC0.2 post-processing scheme will help water managers and users of the forecasting service make better-informed decisions in planning and management of water resources. Highlights. Uncorrected and post-processed streamflow forecasts (using three transformations, namely Log, Log-Sinh, and BC0.2) are evaluated over 300 diverse Australian catchments. Post-processing enhances streamflow forecast reliability, increasing the percentage of catchments with reliable predictions from 50 % to over 90 %. The BC0.2 transformation achieves substantially better forecast sharpness than the Log-Sinh and Log transformations, particularly in dry catchments.

Multi‐temporal Hydrological Residual Error Modeling for Seamless Subseasonal Streamflow Forecasting

“…See Boucher et al (2018) and Hemri (2019) for overviews of QPP, and Demargne et al (2014) and Demeritt et al (2013) for applications in national hydrological forecasting services. QPP has been shown to be effective for correcting biases and underdispersion of probabilistic streamflow forecasts (e.g., Hemri et al, 2015;Verkade et al, 2017;Woldemeskel et al, 2018;Zhao et al, 2011). QPP models can be classified into two broad categories:…”

“…These models account jointly for both rainfall and hydrological uncertainty. Examples of forecast-rainfall-based QPP models include the multisite short term (1-10 days) forecasting model of Engeland and Steinsland (2014) and the monthly/seasonal forecasting model of the Australian Bureau of Meteorology (Tuteja et al, 2011;Woldemeskel et al, 2018); 2. Observed-rainfall-based QPP models are developed and calibrated using hydrological model simulations forced with observed rainfall.…”

A Data Censoring Approach for Predictive Error Modeling of Flow in Ephemeral Rivers

“…For ephemeral rivers, this approach violates the assumptions of normal and symmetrical errors as pointed out by Smith et al (). It has nonetheless been used as a simple “pragmatic” approach in some studies that include ephemeral rivers (McInerney et al, ; Woldemeskel et al, ; Ye et al, ). Predictive uncertainty is generated only with the method used for Condition in section , including when . o‐censored.…”

Generation of Monthly Precipitation Climatologies for Costa Rica Using Irregular Rain-Gauge Observational Networks

“…Box-Cox transformation was individually applied to HSR and HTR datasets for each climatic region at a monthly basis, prior to the application of kriging interpolation. As suggested by Erdin et al [48] and Woldemeskel et al [49], possible values for "λ" were constrained to a minimum value of 0.2 in order to avoid excessive data transformation. Precipitation estimates and kriging variance were subsequently back-transformed to their original units (mm/months).…”