Development of the next-generation air quality prediction system in the Unified Forecast System framework: Enhancing predictability of wildfire air quality impacts

Bulletin of the American Meteorological Society, Год журнала: 2025, Номер unknown

Опубликована: Янв. 15, 2025

Abstract The National Oceanic and Atmospheric Administration (NOAA) has developed an advanced regional air quality prediction system (AQPS) within the Unified Forecast System (UFS) framework to improve representations of wildfire emissions their impacts on predictions. This innovative integrates Environmental Protection Agency’s (EPA) Community Multiscale Air Quality (CMAQ) model as a column chemistry with UFS-based atmospheric model, operating in online mode. calculation gas particulate relies satellite-derived fire products, high-resolution Regional Hourly Advanced Baseline Imager (ABI) Visible Infrared Imaging Radiometer Suite (VIIRS) Emissions (RAVE). A period June July 2023 Quebec Canadian wildfires, which severely impacted United States (US), was chosen case study assess predictive capability UFS-AQM system. predictions fine (PM 2.5 ) ozone (O 3 were evaluated against AirNow observations from 15 14, 2023. results indicate substantial improvement PM when compared previous operational forecast. Meanwhile, demonstrates strong ability predicting O exceedance events during dissipation phase wildfire. Furthermore, shows more realistic aerosol optical depth (AOD) forecast satellite retrieval data. Finally, this outlines plan for further advancing comprehensive AQPS at NOAA.

Язык: Английский

---

Remote sensing for wildfire monitoring: Insights into burned area, emissions, and fire dynamics

One Earth, Год журнала: 2024, Номер 7(6), С. 1022 - 1028

Опубликована: Июнь 1, 2024

Remote sensing plays a central role in monitoring wildfires throughout their life cycle, including assessing pre-fire fuel conditions, characterizing active fire locations and emissions, evaluating post-fire effects on vegetation, air quality, climate. This primer examines current remote products used wildfire research, focusing application deriving burned area emissions data tracking the dynamic spread of individual events. We evaluate strengths weaknesses these address key challenges such as generating complete, continuous, consistent long-term data. also explore future opportunities directions technology for characterization management.

Язык: Английский

---

The Canadian Fire Spread Dataset

Scientific Data, Год журнала: 2024, Номер 11(1)

Опубликована: Июль 11, 2024

Abstract Satellite data are effective for mapping wildfires, particularly in remote locations where monitoring is rare. Geolocated fire detections can be used enhanced management and modelling through daily progression mapping. Here we present the Canadian Fire Spread Dataset (CFSDS), encompassing interpolated progressions fires >1,000 ha Canada from 2002–2021, representing day-of-burning 50 environmental covariates every pixel. Day-of-burning was calculated by ordinary kriging of active Moderate Resolution Imaging Spectroradiometer Visible Infrared Radiometer Suite, enabling a substantial improvement coverage resolution over existing datasets. Day burning at each pixel to identify conditions such as weather, derived weather metrics, topography, forest fuels characteristics. This dataset broad range research applications, retrospective analysis spread, benchmark validating statistical or machine-learning models, forecasting effects climate change on activity.

Язык: Английский

---

Forecasting Daily Fire Radiative Energy Using Data Driven Methods and Machine Learning Techniques

Journal of Geophysical Research Atmospheres, Год журнала: 2024, Номер 129(16)

Опубликована: Авг. 24, 2024

Abstract Increasing impacts of wildfires on Western US air quality highlights the need for forecasts smoke emissions based dynamic modeled wildfires. This work utilizes knowledge weather, fuels, topography, and firefighting, combined with machine learning other statistical methods, to generate 1‐ 2‐day fire radiative energy (FRE). The models are trained data covering 2019 2021 evaluated 2020. For 1‐day (2‐day) forecasts, random forest model shows most skill, explaining 48% (25%) variance in observed daily FRE when all available predictors compared 2% (<0%) explained by persistence extreme year also improved skill forecasting day‐to‐day increases decreases FRE, 28% (39%) increase (decrease) days predicted, identified 62% (60%) accuracy. Error tends toward under severe weather. Sensitivity analysis that near‐surface weather latest contribute model. When was subsets training produced agencies (e.g., Canadian or Forest Services), comparable if not better performance achieved (1‐day R 2 = 0.39–0.48, 0.13–0.34). is used compute emissions, so these results demonstrate potential models.

Язык: Английский

---

Optical properties of biomass burning aerosol during the 2021 Oregon fire season: comparison between wild and prescribed fires

Environmental Science Atmospheres, Год журнала: 2023, Номер 3(3), С. 608 - 626

Опубликована: Янв. 1, 2023

The Mt. Bachelor Observatory was frequently impacted by biomass burning smoke in 2021, an extreme forest fire year the state of Oregon.

Язык: Английский

---

Estimating the impact of a 2017 smoke plume on surface climate over northern Canada with a climate model, satellite retrievals, and weather forecasts

Authorea (Authorea), Год журнала: 2023, Номер unknown

Опубликована: Июнь 8, 2023

In August 2017, a smoke plume from wildfires in British Columbia and the Northwest Territories recirculated persisted over northern Canada for two weeks. We compared full-factorial set of NASA Goddard Institute Space Studies ModelE simulations to satellite retrievals aerosol optical depth carbon monoxide, finding that performance was dependent on model configuration, more so choice injection height approach, scheme biomass burning emissions estimates than horizontal winds nudging. particular, with free-tropospheric injection, mass-based high fire NOx led unrealistically depth. Using paired excluded, we estimated 16 days an 850 000 km2 region, decreased planetary boundary layer heights by between 253 m 547 m, downward shortwave radiation 52 Wm-2 172 Wm-2, surface temperature 1.5 oC 4.9 oC, latter spanning independent estimate operational weather forecasts 3.7 cooling. The strongest climate effects were configurations detailed microphysics stronger first indirect effect.

Язык: Английский

---

Systematically tracking the hourly progression of large wildfires using GOES satellite observations

Earth system science data, Год журнала: 2024, Номер 16(3), С. 1395 - 1424

Опубликована: Март 15, 2024

Abstract. In the western United States, prolonged drought, a warming climate, and historical fuel buildup have contributed to larger more intense wildfires as well longer fire seasons. As these costly become common, new tools methods are essential for improving our understanding of evolution fires how extreme weather conditions, including heat waves, windstorms, droughts, varying levels active-fire suppression, influence spread. Here, we develop Geostationary Operational Environmental Satellites (GOES)-Observed Fire Event Representation (GOFER) algorithm derive hourly progression large create product perimeters, lines, spread rates. Using GOES-East GOES-West geostationary satellite detections active fires, test GOFER on 28 in California from 2019 2021. The includes parameter optimizations defining burned-to-unburned boundary correcting parallax effect elevated terrain. We evaluate perimeters using 12 h data Visible Infrared Imaging Radiometer Suite (VIIRS)-derived Data (FEDS) final California's Resource Assessment Program (FRAP). Although GOES imagery used has coarser resolution (2 km at Equator), correspond reasonably those obtained FRAP, with mean Intersection-over-Union (IoU) 0.77, comparison 0.83 between FEDS FRAP; IoU indicates area overlap over union relative reference which 0 is no agreement 1 perfect agreement. fills key temporal gap present other tracking products that rely low-Earth-orbit imagery, where available intervals or ad hoc aircraft overflights. This particularly relevant when spreads rapidly, such maximum rates 5 h−1. Our deriving can be applied across North South America reveals considerable variability diurnal timescales. resulting broad set potential applications, development predictive models improvement atmospheric transport surface smoke estimates. estimates (https://doi.org/10.5281/zenodo.8327264, Liu et al., 2023).

Язык: Английский

---

Mitigating underestimation of fire emissions from the Advanced Himawari Imager: A machine learning and multi-satellite ensemble approach

International Journal of Applied Earth Observation and Geoinformation, Год журнала: 2024, Номер 128, С. 103784 - 103784

Опубликована: Март 22, 2024

The accurate estimation of biomass burning emissions has played a crucial role in air quality and climate forecast modeling. Satellite-based fire radiative power (FRP) proven effective for calculating emissions. However, FRP-based emission estimations East Asia often rely on polar-orbiting satellites owing to the unstable performance Japan Aerospace Exploration Agency Advanced Himawari Imager (JAXA AHI) from poor detection capability unproper FRP retrieval method. To address this, we improve by machine learning based mid-infrared (MIR) radiance method, leveraging superior model developed our previous study. In addition, propose multi-satellite distance-based weighted ensemble Compared traditional MIR methods, learning-based exhibited promising (correlation coefficient: 1, mean bias error: 0.2, absolute percentage 1.9%). integration dramatically mitigated underestimation issues JAXA AHI. was combined with Moderate Resolution Imaging Spectroradiometer create FRP. Comparative assessments using TROPOspheric Monitoring Instrument conventional bottom-up method demonstrated that proposed produced reliable output. Furthermore, impact analysis revealed missing peaks or underestimated burn scars could lead fatally low emissions; however, relatively robust against data its ensemble. By identifying potential problems their estimations, this study provides valuable insights studies.

Язык: Английский

---

Multi-resolution monitoring of the 2023 maui wildfires, implications and needs for satellite-based wildfire disaster monitoring

Science of Remote Sensing, Год журнала: 2024, Номер 10, С. 100142 - 100142

Опубликована: Июнь 6, 2024

The August 2023 wildfires over the island of Maui, Hawaii were one deadliest U.S. wildfire incidents on record with 100 deaths and an estimated $5.5 billion cost. This study documents incidence, extent, characteristics Maui using multi-resolution global satellite fire products, in so doing demonstrates their utility limitations for detailed monitoring, highlights outstanding observation needs monitoring. NASA 500 m Moderate Resolution Imaging Spectroradiometer (MODIS) burned area product is compared PlanetScope 3 areas that mapped a published deep learning algorithm. In addition, all active detections provided by MODIS Terra Aqua satellites Visible Infrared Radiometer Suite (VIIRS) S-NPP NOAA-20 are used to investigate geographic temporal occurrence fires incidence relative areas. diurnal variation radiative power (FRP), available detections, presented examine how energetically burning. analysis undertaken town Lahaina was major population center burned. Satellite first detected 8th early morning (1:45 onwards) western slopes Mt. Haleakalā last 10th (at 2:46) Haleakalā. FRP VIIRS indicate less intensely from beginning end this three day period, nighttime generally more than daytime fires, most burning occurred likely due high fuel load buildings vegetation elsewhere. too coarse map 18 unambiguously at resolution covered 29.60 km2, equivalent about 1.6% Maui. systematically derived products assessment before, during after disaster events such as those experienced future monitoring events, recommendation constellation, discussed.

Язык: Английский

---

Evaluation of an In‐Canopy Wind and Wind Adjustment Factor Model for Wildfire Spread Applications Across Scales

Journal of Advances in Modeling Earth Systems, Год журнала: 2024, Номер 16(7)

Опубликована: Июль 1, 2024

Abstract The representation of vegetative sub‐canopy wind is critical in numerical weather prediction (NWP) models for the determination air‐surface exchange processes heat, momentum, and trace gases. Because relationship between speed fire behaviors, influence canopy on near‐surface prognostic spread used regional NWP models. In practice, at midflame point fires (midflame speed) to determine rate spread. However, speeds from most situ measurements are taken some reference height above flames. Hence, this study develops a modular computationally‐efficient one‐dimensional model set composed adjustment factor (WAF) applications across scales. uses prescribed foliage shape functions represent vertical vegetation profile its impacts three‐dimensional structure horizontal speeds. Results well agree with ground‐based observations average mean absolute bias, root square error coefficients around 0.18 m s −1 , 0.40 0.90, respectively. WAF provides by estimating based canopy, flame characteristics. Various user‐definable options provide flexibility adapt variations characteristics additional complexities associated wildfires. expected improve providing an improved sub‐grid flows any spatial scale.

Язык: Английский

---