Using soil moisture information to better understand and predict wildfire danger

Researchers and Fire Managers Convene

Recognizing these opportunities, more than 200 soil moisture researchers, fire researchers, fire managers, ecologists, and government personnel registered for a one-day online symposium (https://bit.ly/3njq1MY) held on 6 May 2021. The objectives of the symposium were to provide researchers and fire managers an opportunity to connect with others, learn about ongoing research in this area, and discuss ways to move forward with new research and end uses. Invited speakers discussed current trends and status of: (1) wildfire hazard rating; (2) relationships between remotely sensed soil moisture and wildfire; (3) use of soil moisture models for fire danger assessments; and (4) linkages among in situ soil moisture measurements, fuel bed characteristics, and wildfire. Breakout groups and a panel discussion focused on key remaining questions and next steps.

Through these interactions, participants highlighted the need to present the science in a readable and actionable way for fire managers. Increased partnership between researchers and decision-makers was identified as one key strategy to help meet this need, so that soil moisture information can better fit into existing workflows and decision processes. Such partnerships likely require iterative conversations that may be difficult to sustain via typical grant-funded research projects, which tend to be too short lived to bring research into practice. At the same time, recent increases in wildfire-burned areas across the U.S. (Burke et al., 2021) have increased the stress on fire managers and limited the time and energy available for incorporating new research results into fire management processes. New federal investments are necessary to further develop this promising research area and to integrate the knowledge produced into existing fire danger rating systems and associated fire management tools and workflows.

The symposium’s speakers, panelists, and participants also emphasized the need to go beyond observational studies of the strong correlations between soil moisture and wildfire to more in-depth studies that quantify the underlying mechanisms operating in the soil-plant-atmosphere-fire nexus. Such mechanistic studies require intensive, in situ time-series measurements of not only soil moisture, but also fuel moisture contents and fuel loads, properties which typically require time-consuming, destructive, manual sampling of both live and dead vegetation. Further enhancement and geographic expansion of existing fuel moisture databases and development of dynamic fuel load databases may be necessary to support deeper understanding of the mechanisms linking soil moisture and wildfire.

Looking Deeper at Soil Moisture—Wildfire Relationships

The potential influence of soil moisture on wildfire behavior has long been recognized, beginning with the Keetch–Byram Drought Index (Keetch & Byram, 1968), a soil moisture surrogate that continues to be widely used by wildfire managers and scientists today. But for decades, the limited availability of soil moisture observations with sufficient duration and spatial extent severely restricted progress towards understanding soil moisture—wildfire relationships. This has begun to change in the current era of new sources of soil moisture measurements, which started with the advent of large-scale in situ soil moisture monitoring networks in the late 1990s and continued with the launch shortly thereafter of satellite missions capable of remotely sensing soil moisture. When the resulting data records reached sufficient duration, we began to get our first solid evidence for the strong relationships between wildfire and soil moisture measured via remote sensing (Chen et al., 2013) and in-situ networks (Krueger et al., 2015).

Since then, we have entered a phase of accelerating growth in our understanding of soil moisture—wildfire relationships and of the mechanisms driving those relationships. We now know that in situ soil moisture relates differently to wildfire probability during the growing season than during the dormant season (Krueger et al., 2016) and that in situ soil moisture measurements can better predict large growing-season wildfires than the Keetch–Byram Drought Index (Krueger et al., 2017). We also learned how soil moisture observations can provide an effective indicator of fuel moisture content and curing rate for grassland vegetation (Sharma et al., 2020) and dead fuel moisture content in Sierra Nevada forests (Rakhmatulina et al., 2021). In situ soil moisture measurements have also proven valuable for improving predictions of grassland fuel loads (Krueger et al., 2021), which is significant given the fact that grasslands account for a large portion of the area burned worldwide each year.

Underlying the dynamic soil moisture values are relatively static but influential soil properties. These properties provide a natural integrator of precipitation variability (Chikamoto et al., 2015), and thus detailed soil-mapping approaches have also shown value for fire prediction (Levi & Bestelmeyer, 2016, 2018). In addition, there is a rapidly expanding body of knowledge resulting from studies linking remotely sensed soil moisture to wildfire. Those studies have shown how remotely sensed soil moisture can be used to estimate fuel moisture content and to predict wildfire extent from regional to global scales (Chaparro et al., 2016; Jensen et al., 2018; Lu & Wei, 2021; O et al., 2020; Rigden et al., 2020; Thomas Ambadan et al., 2020). Taken together, there are strong and growing lines of evidence that soil moisture information can help us better understand and predict wildfire danger.

Next Steps

Towards that end, some symposium participants expressed interest in forming working groups to move the research forward. Already one such group has formed and is working on an interdisciplinary review paper to summarize the rapidly growing body of research, broaden the community of researchers aware of and engaged in this line of research, and make a convincing case for more widespread use of soil moisture information in operational fire danger rating systems. Symposium participants also expressed interest in a follow-up meeting to build momentum and sustain progress. Currently, the U.S. Forest Service is organizing one such meeting to be held in spring 2022 in collaboration with USDA-ARS. That meeting is intended as a step toward developing a framework for improved forest soil moisture monitoring across the U.S. Two potential topics are the enhancement of monitoring infrastructure in the nation’s forests and grasslands and how to best synthesize that data into a dashboard to provide quick interpretations for land managers.

Another key next step discussed by symposium participants is to effectively integrate and leverage the increasingly diverse sources of soil moisture information from land surface models, in situ networks, and satellites. Each of these information sources has associated strengths and limitations, and objective methods for blending soil moisture estimates from these sources are still being developed (Zhang et al., 2021). There is a clear need to evaluate how such blended products can be effectively used to estimate fuel moisture, fuel load, and wildfire danger. Meanwhile, the development of soil moisture models designed specifically for application in wildfire danger rating systems is another key area in need of more research.