The Hot-Dry-Windy Index (HDWI) is a fire weather tool used by wildfire forecasters and fire managers across the United States. It judges how much the atmosphere itself, apart from the fuel on the ground or the shape of the land, is likely to make a wildfire harder to control on a given day. It does this by combining three ingredients, how hot the air is, how dry it is, and how hard the wind is blowing, into a single number. That underlying number is usually written HDW (Hot-Dry-Windy). HDWI is the name most often used for the full real-time tool and website the U.S. Forest Service maintains around it. That tool adds a 30-year climate record, day-by-day forecasts, and a library of real-fire case studies on top of the raw number itself. This article covers both: how HDW is calculated, and how the broader HDWI tool presents and interprets it.
What HDW measures
HDW is calculated as the highest wind speed multiplied by the highest vapor pressure deficit found anywhere in roughly the lowest 500 meters (about 1,600 feet) of the atmosphere above a given location (Srock et al., 2018). Vapor pressure deficit is a measure of how much drier the air is than it would be if it were fully saturated with moisture, essentially a measure of the air’s thirst for water. The index’s developers chose vapor pressure deficit over the more familiar relative humidity because relative humidity can hide real differences in dryness at different temperatures. Warm air can hold much more moisture than cold air, so two locations with the same relative humidity reading can differ sharply in how quickly they can dry out vegetation, while vapor pressure deficit captures that difference directly (Srock et al., 2018).
Multiplying wind and dryness together, rather than looking at either one alone, reflects how the two work together in real fire behavior: dry air draws moisture out of vegetation and makes it easier to ignite, and wind then feeds a fire more oxygen and pushes flames and embers into new fuel. Taking the maximum of each variable, rather than an average, is meant to capture the single worst combination of conditions likely to affect a fire during the day, since that peak is often what matters most for fire spread.
The HDWI tool
The HDWI website (hdwindex.fs2c.usda.gov) is maintained by the U.S. Forest Service in partnership with St. Cloud State University, and offers several things beyond a single HDW number: forecasts of HDW several days out built from the Global Ensemble Forecast System (GEFS), a National Weather Service ensemble model that runs the same forecast multiple times with slightly different starting conditions to gauge confidence; a 30-year historical climatology built from a different dataset, the Climate Forecast System Reanalysis (CFSR); a rolling 30-day archive; separate products for Alaska; and a set of case studies pairing HDW’s behavior with real historical wildfires (USDA Forest Service, HDWI).
Because the raw HDW number by itself does not say much without a sense of what is normal for a given place and time of year, the HDWI tool shades its forecast maps by percentile against that 30-year CFSR climatology rather than by the raw hPa·m/s value alone. The tool does not discard the raw number, forecast charts still plot it directly, but overlays percentile-based color shading as context, since what counts as a dangerously high HDW value in a humid coastal climate can be an ordinary summer afternoon in the desert Southwest.
Development and history
HDW was introduced in a 2018 study by Alan F. Srock of St. Cloud State University together with Joseph J. Charney, Brian E. Potter, and Scott L. Goodrick of the U.S. Forest Service’s research stations (Srock et al., 2018). It was developed as a modern alternative to the Haines Index, an older fire-weather index dating to the 1980s that measured atmospheric instability. The Haines Index included no wind term at all, a real gap, since wind is one of the most direct drivers of how fast a fire spreads. Testing HDW against four historical wildfires, the study’s authors found it tracked the day of most active fire behavior more consistently than the Haines Index did (Srock et al., 2018).
The Haines Index has since been formally phased out. The National Wildfire Coordinating Group’s Fire Weather Subcommittee concluded it was “unsupported by research in predicting large fire growth” and recommended retiring it for the 2025 fire season, replacing it with a different metric called Mixing Height, which instead measures how freely a column of rising smoke and heat can climb through the atmosphere (National Wildfire Coordinating Group). HDW, tracking a different question, wind and dryness rather than vertical instability, has continued in operational use through this transition.
Reading HDW values: percentiles, not a fixed scale
Because HDW values depend heavily on a location’s own climate and the time of year, forecasters are generally advised to judge a given HDW reading against that location’s own historical range for the same calendar period, rather than against one fixed number that is meant to apply everywhere. The same numerical HDW value that would be unremarkable at one location or time of year can represent an extreme reading at another, which is the reasoning behind the HDWI tool’s percentile-based climatology described above.
Limitations, including a complex-terrain caveat
Like any index built from broad-scale weather-model or reanalysis data, HDW and HDWI inherit the limitations of whatever grid of data they are calculated from. A model or reanalysis dataset with widely spaced grid points can only represent the atmosphere at that same coarse scale, smoothing over small but locally important features such as narrow canyons, individual ridgelines, or the specific way wind channels through a mountain pass.
The HDWI project’s own case-studies page states this limitation directly, in its account of the Thirtymile Fire, a 2001 wildfire in Washington State during which four firefighters were killed. The case study explains that the fire’s dangerous behavior on the day of the fatalities was “driven primarily by flows tied to the local terrain,” features the reanalysis data underlying HDWI’s climatology “would be incapable of resolving,” and states plainly that “HDWI (and any other fire-weather index calculated using larger-scale meteorological data) should be interpreted with caution in complex terrain” (USDA Forest Service, HDWI Case Studies). Most of HDWI’s other published case studies cover comparatively flat terrain, coastal chaparral, Great Lakes and boreal forest, coastal pine, and Southern Plains grassland, where HDW cleanly identified the day of greatest fire danger; the Thirtymile Fire stands out among them as the case where mountainous terrain, not a large-scale weather pattern, drove the outcome (USDA Forest Service, HDWI Case Studies).
This caveat, stated by HDWI’s own team, is directly relevant anywhere fire weather is being assessed in mountainous or canyon terrain. The same gap between coarse model resolution and fine local terrain applies wherever that mismatch exists, and the Thirtymile Fire is simply the one case where HDWI’s own developers have said so in writing.
HDW on this site
Sky Islands Climate computes its own HDW value for the region’s fire weather product, drawn from the High-Resolution Rapid Refresh (HRRR) model rather than from HDWI’s CFSR-based climatology or GEFS forecasts. HRRR’s much finer grid spacing represents local terrain considerably better than the coarser data behind HDWI’s own climatology, but the terrain-resolution caveat above still applies in kind, just to a lesser degree, in the mountainous “sky islands” terrain this site covers. See the site’s Fire Weather pages for the current HDW readings and their full methodology notes (Sky Islands Climate, Fire Weather).
Sources
Srock, A.F., Charney, J.J., Potter, B.E., & Goodrick, S.L. “The Hot-Dry-Windy Index: A New Fire Weather Index.” Atmosphere, 9(7), 279, July 19, 2018. https://doi.org/10.3390/atmos9070279
USDA Forest Service. “Hot-Dry-Windy Index (HDWI).” https://hdwindex.fs2c.usda.gov/
USDA Forest Service. “HDWI Case Studies.” https://hdwindex.fs2c.usda.gov/case.html
National Wildfire Coordinating Group. “Replacing Haines Index and Lightning Activity Level.” https://www.nwcg.gov/6mfs/weather-fire-behavior/replacing-haines-index-and-lightning-activity-level