Understanding the impacts of hydrological non-stationarity on runoff projections

This report documents the state of knowledge of the secondary, or indirect, impacts of climate change and prioritises topics for investigation in the MD-WERP climate adaption theme.

Science and research

MD-WERP theme	Climate adaptation
MD-WERP theme 1 project	Climate adaptation foundational science
Research lead	CSIRO
Authors	David Robertson, Hongxing Zheng, Julien Lerat and Francis Chiew
Author contact details	David Robertson
Date of publication	December 2024
Report full title	Understanding the impacts of hydrological non-stationarity on runoff projections
Keywords	hydrological non-stationarity; hydrological analysis; hydrological time series; modelling; rainfall-runoff; run-off projections; catchments; calibration periods; daisi; model sensitivity; global warming; vegetation change; water resources
Summary of output	Hydrological non-stationarity refers to changes in the statistical characteristics of catchment rainfall or runoff or the relationship between the two. The origins of hydrological non-stationarities include global warming, vegetation change, water resources development activities, and the cumulative impact of interactions between changing surface and sub-surface processes. Considerable past research has involved diagnosing non-stationarity in hydrological time series and simulations of rainfall-runoff models. This study provides insights into the likely impacts of hydrologic non-stationarity on runoff projections for the Murray–Darling Basin, through two investigations assessing: the sensitivity of runoff projections to the period used to calibrate conceptual rainfall-runoff models the extent to which an approach to adapt existing hydrological models to better reflect catchment rainfall-runoff process alters the model sensitivity of runoff to changes in rainfall.
Key findings / recommendations	Overall, runoff projections are sensitive to hydrological non-stationarity. Due to the differences in the significance of hydrological non-stationarity across the catchments and the ability of the models to reflect these non-stationarities, the sensitivity of runoff projections to hydrological non-stationarity is spatially variable and model dependent. Projections generated using simpler hydrological models appear to be less sensitive to hydrological non-stationarity, but it is well established that simpler models often are limited in their ability to simulate changes in hydrological processes. Future research should pursue the development of hydrological models that can better represent historical non-stationarities using simplified parameterisations which, based on our analysis should produce robust runoff projections.
Target audience	Water managers, researchers, Australian Government, state governments, local governments, community, water and environmental consultants

Report
Publication title	Published	File type	File size
Understanding the impacts of hydrological non-stationarity on runoff projections	13 Dec 2024	PDF	6.54 MB

Published date: 19 May 2025