Predicting wildfire induced changes to runoff: A review and synthesis of modeling approaches

Abstract

Wildfires elicit a diversity of hydrological changes, impacting processes that drive both water quantity and quality. As wildfires increase in frequency and severity, there is a need to assess the implications for the hydrological response. Wildfire‐related hydrological changes operate at three distinct timescales: the immediate fire aftermath, the recovery phase, and long‐term across multiple cycles of wildfire and regrowth. Different dominant processes operate at each timescale. Consequentially, models used to predict wildfire impacts need an explicit representation of different processes, depending on modeling objectives and wildfire impact timescale. We summarize existing data‐driven, conceptual, and physically based models used to assess wildfire impacts on runoff, identifying the dominant assumptions, process representations, timescales, and key limitations of each model type. Given the substantial observed and projected changes to wildfire regimes and associated hydrological impacts, it is likely that physically based models will become increasingly important. This is due to their capacity both to simulate simultaneous changes to multiple processes, and their use of physical and biological principles to support extrapolation beyond the historical record. Yet benefits of physically based models are moderated by their higher data requirements and lower computational speed. We argue that advances in predicting hydrological impacts from wildfire will come through combining these physically based models with new computationally faster conceptual and reduced‐order models. The aim is to combine the strengths and overcome weaknesses of the different model types, enabling simulations of critical water resources scenarios representing wildfire‐induced changes to runoff. This article is categorized under: Water and Life > Conservation, Management, and Awareness Science of Water > Hydrological Processes Science of Water > Water and Environmental Change

Modeling Post‐Wildfire Hydrologic Response: Review and Future Directions for Applications of Physically Based Distributed Simulation

“…Our suggested model characteristics are not prescriptive; advances in hydrologic modeling after wildfire are not tied exclusively to physically based distributed models. Tremendous strides in prediction could be made by fusion of data‐driven models, lumped conceptual models, and physically based models (Partington et al., 2022).…”

“…The recent scoping review by Partington et al. (2022) covered a broad range of post‐wildfire hydrologic models, including data‐driven and lumped (spatially uniform, process‐averaged) models, and identified the need for long‐term integrated models that span fuels regeneration, wildfire ignition, and the hydrologic consequences.…”

“…A recent review by Lopes et al (2020) revealed important gaps in post-wildfire erosion modeling, including the need for more comparison to field data and uncertainty evaluation. The recent scoping review by Partington et al (2022) covered a broad range of post-wildfire hydrologic models, including data-driven and lumped (spatially uniform, process-averaged) models, and identified the need for long-term integrated models that span fuels regeneration, wildfire ignition, and the hydrologic consequences.…”

Changes in Blue/Green Water Partitioning Under Severe Drought

“…To our knowledge, this is the first published application of RHESSys for an Australian ecosystem, and may open the door to future advances like those achieved with RHESSys in the US, Europe and Asia (Bart et al., 2016; Boisramé et al., 2019; Hanan et al., 2017; Kim et al., 2007; Kim Ji et al., 2018; Ren et al., 2022; Tague et al., 2009; Zierl et al., 2007). Some data processing tools for RHESSys setup have already been adapted to work with Australian data sets (Miles & Band, 2015) and the model's potential role in addressing nationally important research topics such as changing fire risk has been highlighted recently (Partington et al., 2022). The catchment we simulated was relatively humid (aridity index ∼1), but many Australian eucalypt forests are more water‐limited and accurately accounting for ecophysiological processes may be even more important for simulating hydrologic behavior (Garcia et al., 2016).…”

A Nonlinear Model for Evaluating Dynamic Resilience of Water Supply Hydropower Generation‐Environment Conservation Nexus System

“…Thus, the nonlinear behavior of a system emerges. For instance, Partington et al (2022) found that resilience processes of runoff exhibited two significantly different shapes in short-term and long-term postfire conditions, and didn't meet the principle of proportionality. Piggott et al (2015) argued that cumulative effect of multiple disturbances on ecosystem were greater or less than the additive sum of effects produced by the disturbances acting in isolation.…”