Spatiotemporal downscaling and nowcasting of urban land surface temperatures with deep neural networks

TL;DR

This paper presents a deep learning approach combining satellite data to generate high-resolution land surface temperature fields and perform intraday nowcasting, improving spatial and temporal accuracy for urban climate applications.

Contribution

It introduces a novel spatiotemporal downscaling method using a U-Net and a ConvLSTM-based nowcasting model trained on European urban data, achieving high accuracy and operational applicability.

Findings

Achieved RMSE of 1.92°C in downscaling LSTs across European cities.

Outperformed benchmarks in LST nowcasting with RMSEs of 0.57 to 1.15°C.

Validated robust performance against independent MODIS data.

Abstract

Land Surface Temperature (LST) is a key variable for various applications, such as urban climate and ecology studies. Yet, existing satellite-derived LST products provide either high spatial or high temporal resolution, resulting in a fundamental trade-off between the two. To address this trade-off, we combine observations from a geostationary and a polar orbiting satellite and provide LST fields at high spatial and high temporal resolution (1 km at 15-min intervals). We demonstrate their application for intraday forecasting of LSTs. To estimate LST fields at high spatiotemporal resolution, a U-Net model is trained to map LST fields from SEVIRI/MSG (3 km and 15 min resolution) to LST fields from Terra/Aqua MODIS (1 km, 4 overpasses per day) that are collocated in space and time. The presented model has been trained on LSTs across large European cities with a population exceeding 1 million inhabitants, and achieves an RMSE = $1.92${\deg}C and near-zero bias MBE = $0.01${\deg}C on the hold-out test set. As a second step, we present an LST nowcasting model based on ConvLSTM architecture, trained across downscaled LST fields with forecast lead times of 15 to 75 minutes. The nowcasting model outperforms a persistence and a Climatological Rolling Median benchmarks, with RMSEs of $0.57$ to $1.15${\deg}C for the considered lead times and biases ranging from $-0.1$ to $0.14${\deg}C. An additional validation conducted against independent MODIS overpasses confirms robust performance. Our LST forecast model at high spatiotemporal resolution is directly applicable to operational satellite-based LST monitoring.