What does this research mean for the field?

A shallow physics-informed neural network incorporating a temporal smoothness constraint accurately models high-resolution urban land surface temperatures and reveals climate-dependent predictors, such as vegetation driving cooling in humid climates and impervious surfaces dominating in subtropical climates. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The research aims to create a model that accurately predicts urban land surface temperature and identifies key factors influencing temperature variations across cities.

May 31, 2026ISPRS annals of the photogrammetry, remote sensing and spatial information sciences0 citationsOpen Access

Modeling Urban Land Surface Temperature Using Physics-Informed Neural Networks (PINNs)

RGRonak Ghanbari MLMarc Linderman HAHossein Arefi

Key Points

The research aims to create a model that accurately predicts urban land surface temperature and identifies key factors influencing temperature variations across cities.
Developed a physics-informed neural network (PINN) for modeling land surface temperature.
Used inputs from Landsat-8 spectral indices, digital elevation model, and meteorological covariates.
Achieved R² of 0.88 in NYC and 0.91 in Austin for urban land surface temperature mapping.
Achieved a performance of R² = 0.88 (RMSE = 1.2 °C) in NYC and R² = 0.91 (RMSE = 0.9 °C) in Austin.
Identified vegetation signals as dominant for cooling in NYC, while impervious-surface proxies were strong in Austin.
Demonstrated that a shallow PINN effectively captures urban heat patterns suitable for climate-sensitive planning.

Abstract

Abstract. A compact physics-informed neural network (PINN) is developed to (i) quantify city-scale accuracy of 30 m urban land surface temperature (LST) maps, (ii) identify influential predictors, and (iii) contrast climate-dependent patterns between New York City (NYC) (humid to sub-humid) and Austin, Texas (humid subtropical). Inputs combine selected Landsat-8 spectral indices, a digital elevation model, and meteorological covariates. LST targets are retrieved from Landsat-8 thermal band 10 (single-channel), quality-screened, and resampled to 30 m for May–September 2023. The loss combines data mean squared error term with a lightweight temporal smoothness prior implemented as a finite-difference term (Δ𝑇⁄Δ𝑡) on same-pixel pairs to reflect heat storage behaviour and discourage unrealistically rapid day to day changes. On the study pixels (in-sample), performance reaches R² = 0.88 (RMSE = 1.2 °C) in NYC and R² = 0.91 (RMSE = 0.9 °C) in Austin; errors are approximately Gaussian with minimal bias. Feature patterns differ by climate: vegetation-related signals dominate cooling in NYC, whereas shortwave-radiation and impervious-surface proxies (e.g., NDBI/NDISI) are strongest in Austin. These findings show that a shallow PINN with a minimal temporal constraint yields accurate, interpretable LST maps suitable for urban-heat-island assessment and climate-sensitive heat-mitigation planning.

Demander à l'IA

Bookmark

View Full Paper