Leveraging Land Cover Priors for Isoprene Emission Super-Resolution

TL;DR

This work tackles the challenge of downscaling isoprene BVOC emissions from satellite-derived data by introducing a transformer-based SR framework that is guided by land-cover priors. It integrates high-resolution isoprene inventories with drivers from cropland, tree cover, LAI, and climate class maps, using a non-parametric data transformation to stabilize training. Across climate-folded experiments and three evaluation scenarios, the approach with three semantic priors consistently improves SR accuracy, though generalization to unseen regions or climates remains challenging. The study demonstrates substantial potential for enhanced atmospheric modeling and air quality applications, while outlining improvements such as dynamic priors and hybrid architectures to bolster robustness and scalability.

Abstract

Remote sensing plays a crucial role in monitoring Earth's ecosystems, yet satellite-derived data often suffer from limited spatial resolution, restricting their applicability in atmospheric modeling and climate research. In this work, we propose a deep learning-based Super-Resolution (SR) framework that leverages land cover information to enhance the spatial accuracy of Biogenic Volatile Organic Compounds (BVOCs) emissions, with a particular focus on isoprene. Our approach integrates land cover priors as emission drivers, capturing spatial patterns more effectively than traditional methods. We evaluate the model's performance across various climate conditions and analyze statistical correlations between isoprene emissions and key environmental information such as cropland and tree cover data. Additionally, we assess the generalization capabilities of our SR model by applying it to unseen climate zones and geographical regions. Experimental results demonstrate that incorporating land cover data significantly improves emission SR accuracy, particularly in heterogeneous landscapes. This study contributes to atmospheric chemistry and climate modeling by providing a cost-effective, data-driven approach to refining BVOC emission maps. The proposed method enhances the usability of satellite-based emissions data, supporting applications in air quality forecasting, climate impact assessments, and environmental studies.

Figures (16)

• Figure 1: The proposed deployment pipeline of our system for isoprene emission , leveraging additional emission drivers in the process.
• Figure 2: The proposed training pipeline of our system for isoprene emission , leveraging additional emission drivers in the process.
• Figure 3: The study area adopted in this work; the map report a monthly averaged isoprene emission (April 2019) from the $\text{TD-TROPO-010}$ inventory. Emission is reported as $\frac{mol}{cm^{2}s{1}}$.
• Figure 4: The (\ref{['fig:cl_example']}), (\ref{['fig:tc_example']}) percentage maps and information (\ref{['fig:lai_example']}) over the study area from the $\text{LC-ESA}$ and $\text{LAI-TROPO-010}$ inventories, respectively. For the map in (\ref{['fig:lai_example']}), we report a monthly averaged value (April 2019).
• Figure 5: Geographical distribution of the Köppen-Geiger climate classes that are present over the study area. Please refer to Table \ref{['tab:kg_classes_study_area']} for more information regarding the different climate classes.