This study investigates high-frequency mapping of downward shortwave radiation (DSR) at Earth's surface using the Advanced Baseline Imager (ABI) instrument mounted on Geostationary Operational Environmental Satellite-R Series (GOES-R). The existing GOES-R DSR product (DSR ABI) offers hourly temporal resolution and spatial resolution of 0.25 o. To enhance these resolutions, we explore machine learning (ML) for DSR estimation at the native temporal resolution of GOES-R Level-2 Cloud and Moisture Imagery (CMI) product (five minutes) and its native spatial resolution of two-kilometer at nadir. We compared four common ML regression models through the Leave-One-Out Cross-Validation (LOOCV) algorithm for robust model assessment against ground measurements from AmeriFlux and SURFRAD networks. Results show that Gradient Boosting Regression (GBR) achieves the best performance (R² = 0.916, RMSE = 88.05 W m-2) with efficient computation compared to Long Short-Term Memory (LSTM), which exhibited similar performance. DSR estimates from the GBR model (DSR ALIVE) outperform DSR ABI across various temporal resolutions and sky conditions. DSR ALIVE agreement with ground measurements at SURFRAD networks exhibits high accuracy at high temporal resolutions (five-minute intervals) with R² exceeding 0.85 and RMSE=122 W m-2. We conclude that GBR offers a promising approach for high-frequency DSR mapping from GOES-R, enabling improved applications for near-real-time monitoring of terrestrial carbon and water fluxes.

Land surface temperature (LST) is crucial for understanding earth system processes. We expanded the Advanced Baseline Imager Live Imaging of Vegetated Ecosystems (ALIVE) framework to estimate LST in near-real-time for both cloudy and clear sky conditions at a 5-minute resolution. We compared two machine learning models, Long Short-Term Memory (LSTM) networks and Gradient Boosting Regressor (GBR), using top-of-atmosphere (TOA) observations from the Advanced Baseline Imager (ABI) on the GOES-16 satellite against observations from hundreds of measurement locations for a 5-year period. LSTM outperformed, especially at coarser resolutions and under challenging conditions, with a clear sky R² of 0.96 (RMSE 2.31 K) and a cloudy sky R² of 0.83 (RMSE 4.10 K) across CONUS, based on 10-repeat Leave-One-Out Cross-Validation (LOOCV). GBR maintained high accuracy (R² > 0.90) and ran 5.3 times faster, with only a 0.01-0.02 R² drop. Feature importance revealed infrared bands were key in both models, with LSTM adapting dynamically to atmospheric changes, while GBR utilized time information in cloudy conditions. A comparative analysis against the physically based ABILST product showed strong agreement in winter, particularly under clear sky conditions, while also highlighting the challenges of summer LST estimation due to increased thermal variability. This study underscores the strengths and limitations of data-driven models for LST estimation and suggests potential pathways for integrating these approaches to enhance the accuracy and coverage of LST products.

Monitoring Gross Primary Productivity (GPP), the rate at which ecosystems fix atmospheric carbon dioxide, is crucial for understanding global carbon cycling. Remote sensing offers a powerful tool for monitoring GPP using vegetation indices (VIs) derived from visible and near-infrared reflectance (NIRv). While promising, these VIs often suffer from sensitivity to soil background, moisture, and variations in solar and view zenith angle (SZA and VZA). This study investigates the potential of incorporating shortwave infrared (SWIR) reflectance from MODIS and GOES-R advanced baseline imager (ABI) sensors to improve GPP estimation. We evaluated various formulations for creating SWIR-enhanced Near-InfraRed reflectance of Vegetation (sNIRv) by integrating SWIR information into established VIs across 96 Ameriflux research sites. Our findings reveal that sNIRv improves correlation with GPP for ABI data by up to 0.19 on a half-hourly basis for normalized difference vegetation index (NDVI) values below 0.25, with diminishing gains as NDVI values rise. Using MODIS data, sNIRv matches r values of NIRv for NDVI above 0.25, with a slight 0.05 increase for NDVI below 0.25. Analyses using SCOPE model simulations further support the ability of sNIRv to capture fPAR (fractional photosynthetically active radiation), a proxy for GPP, especially for ecosystems with low LAI. Results highlight that sNIRv-based VIs are less sensitive to soil background, SZA, and VZA compared to NIRv. Shapely Additive Explanations (SHAP) value analysis also identifies sNIRv as the best feature for GPP estimation using machine learning modeling across all different land covers, NDVI ranges, and soil water content (SWC) levels.