Arvind Renganathan

Accurately quantifying terrestrial carbon exchange is essential for climate policy and carbon accounting, yet models must generalize to ecosystems underrepresented in sparse eddy covariance observations. Despite this challenge being a natural instance of zero-shot spatial transfer learning for time series regression, no standardized benchmark exists to rigorously evaluate model performance across geographically distinct climate regimes and vegetation types. We introduce CarbonFluxBench, the first benchmark for zero-shot spatial transfer in carbon flux upscaling. CarbonFluxBench comprises over 1.1 million daily observations from 474 flux tower sites globally (2000–2024). It provides: (1) stratified evaluation protocols that explicitly test generalization across unseen vegetation types and climate regimes, separating spatial transfer from temporal autocorrelation; (2) a harmonized set of remote sensing and meteorological features to enable flexible architecture design; and (3) baselines ranging from tree-based methods to domain-generalization architectures. By bridging machine learning methodologies and Earth system science, CarbonFluxBench aims to enable systematic comparison of transfer learning methods, serves as a testbed for regression under distribution shift, and contributes to the next-generation climate modeling efforts.

We present a knowledge‐guided machine learning framework for operational hydrologic forecasting at the catchment scale. Our approach, a Factorized Hierarchical Neural Network (FHNN), has two main components: inverse and forward models. The inverse model uses observed precipitation, temperature, and streamflow data to generate a representation of the current underlying catchment state. The forward model predicts streamflow using the learned catchment state. The FHNN architecture is designed to model multi‐scale processes and capture their interactions, a critical ability for flood modeling. FHNN also improves forecasts based on real‐time data through an inference‐based data integration approach using inverse modeling. FHNN's data integration approach improves forecasts in response to observed data more efficiently than data assimilation methods that require computationally intensive optimization. We compare the FHNN to a leading deep learning alternative (autoregressive LSTM) on the large‐sample CAMELS‐US data set, and operational flood forecast data from the US National Weather Service (NWS). Official NWS flood forecasts are generated by expert human forecasters using a physics‐based model, in a human‐in‐the‐loop process. Thus, we assess the flood forecast ability of FHNN by directly comparing its performance against these NWS expert‐derived forecasts. The human forecaster creates a more accurate forecast within the first 12–18 hr of a forecast's issuance, but FHNN has significantly better predictions thereafter. This research lays the groundwork for leveraging the predictive performance of AI‐based models with the expertise in forecasting agencies to produce better river forecasts.

We present a knowledge-guided machine learning framework for operational hydrologic forecasting at the catchment scale. Our approach, a Factorized Hierarchical Neural Network (FHNN), has two main components: inverse and forward models. The inverse model uses observed precipitation, temperature, and streamflow data to generate a representation of the current underlying catchment state. The forward model predicts streamflow using the learned catchment state. The FHNN architecture is designed to model multi-scale processes and capture their interactions, a critical ability for flood modeling. FHNN also improves forecasts based on real-time data through an inference-based data integration approach using inverse modeling. FHNN’s data integration approach improves forecasts in response to observed data more efficiently than data assimilation methods that require computationally intensive optimization. To show the performance of FHNN, we compare the FHNN to a leading deep learning alternative (autoregressive LSTM) on the large-sample CAMELS-US dataset, and operational flood forecast data from the US National Weather Service (NWS). Official NWS flood forecasts are generated by expert human forecasters using a physics-based model, in a human-in-the-loop process. Thus, we assess the flood forecast ability of FHNN by directly comparing its performance against these NWS expert-derived forecasts. The human forecaster creates a more accurate forecast within the first 18 hours of a forecast’s issuance, but FHNN has significantly better predictions thereafter. This research lays the groundwork for leveraging the predictive performance of AI-based models with the expertise in forecasting agencies to produce better river forecasts.

Accurate streamflow prediction is crucial to understand climate impacts on water resources and develop effective adaption strategies. A global long short-term memory (LSTM) model, using data from multiple basins, can enhance streamflow prediction, yet acquiring detailed basin attributes remains a challenge. To overcome this, we introduce the Geo-vision transformer (ViT)-LSTM model, a novel approach that enriches LSTM predictions by integrating basin attributes derived from remote sensing with a ViT architecture. Applied to 531 basins across the Contiguous United States, our method demonstrated superior prediction accuracy in both temporal and spatiotemporal extrapolation scenarios. Geo-ViT-LSTM marks a significant advancement in land surface modeling, providing a more comprehensive and effective tool for better understanding the environment responses to climate change.

Streamflow prediction is a long‐standing hydrologic problem. Development of models for streamflow prediction often requires incorporation of catchment physical descriptors to characterize the associated complex hydrological processes. Across different scales of catchments, these physical descriptors also allow models to extrapolate hydrologic information from one catchment to others, a process referred to as “regionalization”. Recently, in gauged basin scenarios, deep learning models have been shown to achieve state of the art regionalization performance by building a global hydrologic model. These models predict streamflow given catchment physical descriptors and weather forcing data. However, these physical descriptors are by their nature uncertain, sometimes incomplete, or even unavailable in certain cases, which limits the applicability of this approach. In this paper, we show that by assigning a vector of random values as a surrogate for catchment physical descriptors, we can achieve robust regionalization performance under a gauged prediction scenario. Our results show that the deep learning model using our proposed random vector approach achieves a predictive performance comparable to that of the model using actual physical descriptors. The random vector approach yields robust performance under different data sparsity scenarios and deep learning model selections. Furthermore, based on the use of random vectors, high‐dimensional characterization improves regionalization performance in gauged basin scenario when physical descriptors are uncertain, or insufficient.

Streamflow prediction is a long-standing hydrologic problem. Development of models for streamflow prediction often requires incorporation of catchment physical descriptors to characterize the associated complex hydrological processes. Across different scales of catchments, these physical descriptors also allow models to extrapolate hydrologic information from one catchment to others, a process referred to as “regionalization”. Recently, in gauged basin scenarios, deep learning models have been shown to achieve state of the art regionalization performance by building a global hydrologic model. These models predict streamflow given catchment physical descriptors and weather forcing data. However, these physical descriptors are by their nature uncertain, sometimes incomplete, or even unavailable in certain cases, which limits the applicability of this approach. In this paper, we show that by assigning a vector of random values as a surrogate for catchment physical descriptors, we can achieve robust regionalization performance under a gauged prediction scenario. Our results show that the deep learning model using our proposed random vector approach achieves a predictive performance comparable to that of the model using actual physical descriptors. The random vector approach yields robust performance under different data sparsity scenarios and deep learning model selections. Furthermore, based on the use of random vectors, high-dimensional characterization improves regionalization performance in gauged basin scenario when physical descriptors are uncertain, or insufficient.

Streamflow prediction is a long-standing hydrologic problem. Development of models for streamflow prediction often requires incorporation of catchment physical descriptors to characterize the associated complex hydrological processes. Across different scales of catchments, these physical descriptors also allow models to extrapolate hydrologic information from one catchment to others, a process referred to as “regionalization”. Recently, in gauged basin scenarios, deep learning models have been shown to achieve state of the art regionalization performance by building a global hydrologic model. These models predict streamflow given catchment physical descriptors and weather forcing data. However, these physical descriptors are by their nature uncertain, sometimes incomplete, or even unavailable in certain cases, which limits the applicability of this approach. In this paper, we show that by assigning a vector of random values as a surrogate for catchment physical descriptors, we can achieve robust regionalization performance under a gauged prediction scenario. Our results show that the deep learning model using our proposed random vector approach achieves a predictive performance comparable to that of the model using actual physical descriptors. The random vector approach yields robust performance under different data sparsity scenarios and deep learning model selections. Furthermore, based on the use of random vectors, high-dimensional characterization identifies the uniqueness of catchments, thereby improving regionalization performance in gauged basin scenario when physical descriptors are uncertain, or insufficient.

Streamflow prediction is a long-standing hydrologic problem. Development of models for streamflow prediction often requires incorporation of catchment physical descriptors to characterize the associated complex hydrological processes. Across different scales of catchments, these physical descriptors also allow models to extrapolate hydrologic information from one catchment to others, a process referred to as “regionalization”. Recently, in gauged basin scenario, deep learning models have been shown to achieve state of the art regionalization performance by building a global hydrologic model. These models predict streamflow given catchment physical descriptors and weather forcing data. However, these physical descriptors are by their nature uncertain, sometimes incomplete, or even unavailable in certain cases, which limits the applicability of this approach. In this paper, we show that by assigning a vector of random values as a surrogate for catchment physical descriptors, we can achieve robust regionalization performance under gauged prediction scenario. Our results show that the deep learning model using our proposed random vector approach achieves a comparable and even marginally better predictive performance than the model using actual physical descriptors. The random vector approach yields robust performance under different data sparsity scenarios and deep learning model selections. Furthermore, our proposed random vector approach provides higher performance for regional modeling when physical descriptors are uncertain, or insufficient.