Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 26 of the 27 references it cites were found in a registry or an index; 3 of the 18 sources its claims cite were checked for support.

Extreme heat prediction through deep learning and explainable AI

Open access

Abstract

Extreme heat waves are causing widespread concern for comprehensive studies on their ecological and societal implications. With the ongoing rise in global temperatures, precise forecasting of heatwaves becomes increasingly crucial for proactive planning and ensuring safety. This study investigates the efficacy of deep learning (DL) models, including Artificial Neural Network (ANN), Conolutional Neural Network (CNN) and Long-Short Term Memory (LSTM), using five years of meteorological data from Pakistan Meteorological Department (PMD), by integrating Explainable AI (XAI) techniques to enhance the interpretability of models. Although Weather forecasting has advanced in predicting sunshine, rain, clouds, and general weather patterns, the study of extreme heat, particularly using advanced computer models, remains largely unexplored, overlooking this gap risks significant disruptions in daily life. Our study addresses this gap by collecting five years of weather dataset and developing a comprehensive framework integrating DL and XAI models for extreme heat prediction. Key variables such as temperature, pressure, humidity, wind, and precipitation are examined. Our findings demonstrate that the LSTM model outperforms others with a lead time of 1–3 days and minimal error metrics, achieving an accuracy of 96.2%. Through the utilization of SHAP and LIME XAI methods, we elucidate the significance of humidity and maximum temperature in accurately predicting extreme heat events. Overall,…

Overview

CoChat Summary

## TL;DR

This paper proposes a deep learning framework for predicting extreme heat in Lahore using five years of PMD weather data (2018–2022) and then explains the predictions with SHAP and LIME. The LSTM model performed best, reaching 96.2% accuracy, and the XAI analysis identified humidity and maximum temperature as the most influential features.

## Key findings

  • The study used 1,450 data instances from the XWeather dataset, split 80% for training and 20% for testing.
  • Among ANN, CNN, and LSTM, the LSTM model was the most effective, with the highest accuracy of 96.2% and the lowest error metrics reported in the paper.
  • A regional extreme-heat threshold was set at the 95th percentile of maximum temperature, corresponding to 41 °C, with a lead time of 1–3 days.
  • SHAP and LIME both indicated that humidity and maximum temperature were the most influential variables for predicting extreme heat events.

## Methods

  • The paper used a regional meteorological dataset (“XWeather”) for Lahore, Pakistan, collected from PMD over five years (2018–2022), with 12 parameters including pressure, humidity, temperature, precipitation, and wind variables.
  • Preprocessing included handling missing values with column averages, label encoding for categorical variables such as wind direction, and min-max numerical scaling.
  • Three deep learning models were trained and compared: ANN, CNN, and LSTM, using Python with Keras/TensorFlow; model performance was evaluated with accuracy, RMSE, MAE, and MSE, and SHAP/LIME were applied for interpretability.

## Limitations

  • The paper states that future work should include data from desert regions like Al-Ahsa in Saudi Arabia, suggesting the current study is limited to Lahore and may not generalize to other climates.
  • The paper also suggests that larger datasets and additional data sources such as satellite imagery and social media could improve performance, implying the current dataset and feature set are limited.
  • Evident from the text: the study reports only one regional dataset and one city-level application, so broader external validation is not shown.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0316367

Cites

26 of 26 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig 1 Proposed methodology for predicting extreme heat events.
Fig 1 Proposed methodology for predicting extreme heat events.CC BY · PubMed Central
Fig 2 Monthly average maximum temperature of study region over the past 5 years.
Fig 2 Monthly average maximum temperature of study region over the past 5 years.CC BY · PubMed Central
Fig 3 Correlation Heatmap of XWeather dataset parameters.
Fig 3 Correlation Heatmap of XWeather dataset parameters.CC BY · PubMed Central
Fig 4 ANN structure for extreme heat prediction.
Fig 4 ANN structure for extreme heat prediction.CC BY · PubMed Central
Fig 5 CNN structure for extreme heat prediction.
Fig 5 CNN structure for extreme heat prediction.CC BY · PubMed Central
Fig 6 Detailed gate structure of LSTM architecture reveals intricate insights into long short-term memory networks.
Fig 6 Detailed gate structure of LSTM architecture reveals intricate insights into long short-term memory networks.CC BY · PubMed Central
Fig 7 LSTM structure for extreme heat prediction.
Fig 7 LSTM structure for extreme heat prediction.CC BY · PubMed Central
Fig 8 Training vs. validation loss of ANN model.
Fig 8 Training vs. validation loss of ANN model.CC BY · PubMed Central
Fig 9 Training VS. validation accuracy of ANN model.
Fig 9 Training VS. validation accuracy of ANN model.CC BY · PubMed Central
Fig 10 Training vs. validation loss of CNN model.
Fig 10 Training vs. validation loss of CNN model.CC BY · PubMed Central
Fig 11 Training vs. validation accuracy of CNN model.
Fig 11 Training vs. validation accuracy of CNN model.CC BY · PubMed Central
Fig 12 Training vs. validation loss of LSTM model.
Fig 12 Training vs. validation loss of LSTM model.CC BY · PubMed Central
Fig 13 Training vs. validation accuracy of LSTM model.
Fig 13 Training vs. validation accuracy of LSTM model.CC BY · PubMed Central
Fig 14 Performance comparison of deep learning models.
Fig 14 Performance comparison of deep learning models.CC BY · PubMed Central
Fig 15 Comparison of actual and predicted heat events: a deep learning perspective.
Fig 15 Comparison of actual and predicted heat events: a deep learning perspective.CC BY · PubMed Central
Fig 16 Understanding the SHAP XAI process.
Fig 16 Understanding the SHAP XAI process.CC BY · PubMed Central
Fig 17 Visualizing XAI SHAP model results for deep learning insight.
Fig 17 Visualizing XAI SHAP model results for deep learning insight.CC BY · PubMed Central
Fig 18 Visual representation of XAI LIME model results.
Fig 18 Visual representation of XAI LIME model results.CC BY · PubMed Central

Verification

Claims the paper makes

1 Valid · 2 Not valid · 7 Not checked · 20 Own results and interpretations

7 more claims

Create a free account to see all 30 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 6
  • It may state two separate claims as one. 14
  • The quote only partly states this claim. 4

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 27 references

Research works printed but not linked: 8

  1. S Sajjad, 2015, The long-term variability in minimum and maximum temperature trends and heat island of Lahore city, Pakistan. Sci Int., 27, 1321
  2. A Perera, 2020, Recent climatic trends in Trinidad and Tobago, West Indies. Asia-Pac J Sci Technol., 25, 1
  3. C O’Brien-Delpesh, 2022, Preserving fragile ecosystems from oil spills – an environmental sensitivity assessment of the east coast of Trinidad, Ocean Coastal Manag, 230, 106374, 10.1016/j.ocecoaman.2022.106374 · DOI · Paper page
  4. A Perera, 2020, Recent climatic trends in Trinidad and Tobago, West Indies. Asia-Pac J Sci Technol., 25, 1
  5. RSS Ranasinghe, 2024, Eco-friendly mix design of slag-ash-based geopolymer concrete using explainable deep learning, Results Eng, 23, 102503, 10.1016/j.rineng.2024.102503 · DOI · Paper page
  6. ND Wimalagunarathna, 2024, Effect of endogenous and anthropogenic factors on the alkalinisation and salinisation of freshwater in United States by using explainable machine learning, Case Stud Chem Environ Eng, 10, 100919, 10.1016/j.cscee.2024.100919 · DOI · Paper page
  7. UAKK Perera, 2024, A new frontier in streamflow modeling in ungauged basins with sparse data: a modified generative adversarial network with explainable AI, Results Eng, 21, 101920, 10.1016/j.rineng.2024.101920 · DOI · Paper page
  8. A Dikshit, 2021, Interpretable and explainable AI (XAI) model for spatial drought prediction, Sci Total Environ, 801, 149797, 10.1016/j.scitotenv.2021.149797 · DOI · Paper page

Cited by

See all 8

Metadata

Authors
Shafiq, F., Zafar, A., Ghani Khan, M. U., Iqbal, S., Albesher, A. S., & Asghar, M. N.
Affiliations
  • University of Engineering and Technology Lahore (PK)
  • King Faisal University (SA)
Corresponding author
Iqbal, Sajid
Venue
PLoS ONE
Published
2025-03-20
Volume
20
Issue
3
Pages
e0316367
ISSN
1932-6203 (ISSN-L)
OpenAlex
W4408659870
Cited by
8
References
27
Citations by year
2026: 6 · 2025: 2
FWCI
2.20
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
Deanship of Scientific Research, King Faisal University (GrantA468)
Topics
  • Physical Sciences › Environmental Science › Environmental Engineering › Hydrological Forecasting Using AI
  • Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Meteorological Phenomena and Simulations
  • Physical Sciences › Environmental Science › Global and Planetary Change › Climate variability and models
Keywords
heatwave prediction, explainable AI, SHAP, lime, LSTM, artificial neural networks, convolutional neural networks, meteorological observations, humidity, maximum temperature
MeSH terms
Deep Learning, Forecasting, Humans, Pakistan, Weather, Neural Networks, Computer, Extreme Heat

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Extreme heat prediction through deep learning and explainable AI. https://app.cochat.ai/papers/10.1371/journal.pone.0316367
The paper
Shafiq, F., Zafar, A., Ghani Khan, M. U., Iqbal, S., Albesher, A. S., & Asghar, M. N. (2025). Extreme heat prediction through deep learning and explainable AI. PLoS ONE. https://doi.org/10.1371/journal.pone.0316367

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.