PyroAdapt: Adapting Wildfire Prediction under Spatial Heterogeneity and Temporal Shift

Hugging Face Daily Papers Papers
wildfire-prediction spatial-heterogeneity temporal-shift machine-learning adaptation-framework risk-ranking environmental-ai

Summary

PyroAdapt proposes a pretrain-retrieve-rank framework to adapt wildfire prediction models for spatial heterogeneity and temporal distribution shifts, improving detection accuracy and prioritizing fire-prone locations under budget constraints.

Prediction of wildfire occurrence is a rare-event problem compounded by spatial heterogeneity and temporal distribution shift, as fire occurrences are vastly outnumbered by non-occurrences, and predictor--fire relationship varies across space and time. Models trained on historical fire data may perform poorly under new conditions and require adaptation to the target distribution before operational use. We propose PyroAdapt, a pretrain--retrieve--rank framework that adapts a pretrained model to target conditions by retrieving historical locations with similar conditions and fine-tuning on the retrievals through risk ranking. For spatial adaptation, we condition risk on terrain, ecoregion embeddings, and fire rates, accounting for spatial context in the retrieval, and learn risk ordering from same-day fire--nonfire cell pairs. We compare direct ranking, residual pairwise DPO (RDPO), and selective ranking through a unified score-gap formulation that characterizes their gradient allocation. Over California (discretized into 666 0.25x0.25 grid cells), these objectives raise daily average precision from 21.62% for continued focal fine-tuning to 24.35--24.57%, and Top5% recall from 18.70% to 22.11--22.79%. Under a fixed daily detection budget of 34 cells (5% area), selective ranking captures 344 additional positive cell--days. For fires in the top 5%/10%/20% of dry matter consumption, selective ranking raises recall by 39.70/28.18/20.50 percentage points, respectively. Furthermore, rolling evaluations over Yosemite show that the gains from ranking persist under temporal distribution shift. Together, these results show that PyroAdapt prioritizes the most fire-prone locations under a daily budget constraint and detects more extreme fire events.
Original Article
View Cached Full Text

Cached at: 09/29/26, 04:08 AM

Paper page - PyroAdapt: Adapting Wildfire Prediction under Spatial Heterogeneity and Temporal Shift

Source: https://huggingface.co/papers/2605.12435

Abstract

Predictionofwildfireoccurrenceisarare-eventproblemcompoundedbyspatialheterogeneityandtemporaldistributionshift,asfireoccurrencesarevastlyoutnumberedbynon-occurrences,andpredictor--firerelationshipvariesacrossspaceandtime.Modelstrainedonhistoricalfiredatamayperformpoorlyundernewconditionsandrequireadaptationtothetargetdistributionbeforeoperationaluse.WeproposePyroAdapt,apretrain--retrieve--rankframeworkthatadaptsapretrainedmodeltotargetconditionsbyretrievinghistoricallocationswithsimilarconditionsandfine-tuningontheretrievalsthroughriskranking.Forspatialadaptation,weconditionriskonterrain,ecoregionembeddings,andfirerates,accountingforspatialcontextintheretrieval,andlearnriskorderingfromsame-dayfire--nonfirecellpairs.Wecomparedirectranking,residualpairwiseDPO(RDPO),andselectiverankingthroughaunifiedscore-gapformulationthatcharacterizestheirgradientallocation.OverCalifornia(discretizedinto6660.25x0.25gridcells),theseobjectivesraisedailyaverageprecisionfrom21.62%forcontinuedfocalfine-tuningto24.35--24.57%,andTop5%recallfrom18.70%to22.11--22.79%.Underafixeddailydetectionbudgetof34cells(5%area),selectiverankingcaptures344additionalpositivecell--days.Forfiresinthetop5%/10%/20%ofdrymatterconsumption,selectiverankingraisesrecallby39.70/28.18/20.50percentagepoints,respectively.Furthermore,rollingevaluationsoverYosemiteshowthatthegainsfromrankingpersistundertemporaldistributionshift.Together,theseresultsshowthatPyroAdaptprioritizesthemostfire-pronelocationsunderadailybudgetconstraintanddetectsmoreextremefireevents.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.12435

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.12435 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.12435 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.12435 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Domain-Adaptive Climate Downscaling Under Temporal Distribution Shift

arXiv cs.LG

This paper investigates temporal out-of-distribution shift in deep-learning-based climate downscaling and proposes a domain-adaptive framework that combines supervised reconstruction with domain alignment to improve high-resolution climate projections under non-stationary conditions.