Assessing and predicting the impacts of wildfires on water quality in Portugal: Project FRISCO

Wildfires change vegetation cover and soil properties, and create a layer of highly mobile ash. Enhanced runoff generation induced by the wildfire can transfer can transfer the ash to streams and water bodies, potentially causing contamination problems and imposing limitations on human uses. This is the focus of project FRISCO: Managing fire-induced risks of water quality contamination, an ongoing applied research project funded by the Portuguese government (PCIF/MPG/0044/2018). The project brings together scientists and water managers to develop tools to better assess post-fire water contamination risks, and identify the best risk mitigation options for water managers working in fire-prone watersheds.

This presentation will focus on the work to identify the main impacts of fires on surface water quality in Portugal based on an analysis of existing water quality data, and to examine the drivers for water quality deterioration. The Portuguese water quality database was examined for sampling stations in watersheds without significant upstream modifications, with data for at least 4 years before and after wildfires of at least 100 ha burned area. In the period 2000-2020, 28 sampling points in rivers and 15 points in 10 reservoirs were found with these characteristics, with monthly data for 6 parameters which are direct and indirect indicators of fire impacts in water quality. The datasets were assessed using Change-Point Analysis to determine the occurrence of breakpoints in water quality following the fire occurrence.

The results show that most fires led to changes in indirect indicators of fire impacts, including significant decrease in dissolved oxygen and pH levels in water, and significant increase of water conductivity. Moreover, changes to direct indicators of fire impacts occurred mostly in reservoir sampling stations, with significant increase in suspended solids, chemical oxygen demand and nitrates concentrations. An in-depth analysis of some stations indicates that the monthly sampling frequency might not be sufficient to capture fast changes associated with large post-fire rainfall events, which nevertheless have impacts on downstream reservoirs. This highlights the importance of finding proxies for water quality impacts which can either be monitored continuously or which can highlight fire impacts despite a monthly sampling frequency.

The work to identify drivers for these changes is ongoing. Special effort is being put in mapping the characteristics of the fires which occurred in the watersheds of the sampling points, including: (i) fire severity, assessed from satellite imagery using the difference Normalized Burn Ratio dNBR index; and (ii) hydrological connectivity between burned areas and the stream network, assessed using the Index of Connectivity by Borselli and Cavalli.

There is also ongoing communication with a group of water managers which operate in fire-prone watersheds, to (i) assess the resilience of the water intake and treatment operations, as well as the capacity to respond to different water qualities, (i) identify the levels of water quality impacts that can be classified as being of concern, and (iii) discuss potential risk mitigation options, including interventions before fire occurrence, emergency post-fire interventions, and changes to water quality treatment.