In late 2023, the Amazon River Basin experienced its most extreme drought, putting its population and ecosystem at major risks. Gauges that were still functioning measured the lowest river water levels (RWL) on record. Here, satellite observations, including Surface Water Ocean Topography (SWOT), reveal the spread and timing of extremely low RWL across the entire river system. Nadir altimeter observations show that the 2023 minimum RWL in Central Amazon were more than 3 m below their annual average, representing two to three times its mean variability. Additionally, SWOT enables the first basin-scale characterization of the reduction in RWL with a spatial resolution of 200 m and how it propagates with time. Large-scale evaluation with gauges suggest that SWOT outperforms classical altimetry in estimating RWL, even if observed biases need further investigations. SWOT offers a new opportunity to understand hydroclimatic extremes and their broad impacts on the environment of the Amazon.

To advance the discovery of scale-relevant hydrological laws while better exploiting massive multi-source data, merging machine learning into process-based modeling is compelling, as recently demonstrated in lumped hydrological modeling. This article introduces MLPM-PR, a new and powerful framework standing for Multiscale spatially distributed Learnable Physical Modeling and learnable Parameter Regionalization with data assimilation. MLPM-PR crucially builds on a differentiable model that couples (i) two neural networks for processes learning and parameters regionalization, (ii) grid-based conceptual hydrological operators, and (iii) a simple kinematic wave routing. The approach is tested on a challenging flash flood-prone multi-catchment modeling setup at high spatio-temporal resolution (1km, 1h). Discharge prediction performances highlight the accuracy and robustness of MLPM-PR compared to classical approaches in both spatial and temporal validation. The physical interpretability of spatially distributed parameters and internal states shows the nuanced behavior of the hybrid model and its adaptability to diverse hydrological responses.

The forthcoming Surface Water and Ocean Topography (SWOT) mission will vastly expand measurements of global rivers, providing critical new datasets for both gaged and ungaged basins. SWOT discharge products will provide discharge for all river reaches wider than 100 m, but at lower accuracy and temporal resolution than what is possible in situ. In this paper, we describe how SWOT discharge produced and archived by the US and French space agencies will be computed from measurements of river water surface elevation, width, and slope and ancillary data, along with expected discharge accuracy. We present here for the first time a complete estimate of SWOT discharge uncertainty budget, with separate terms for random (standard error) and systematic (bias) uncertainty components in river discharge timeseries. We expect that discharge uncertainty will be less than 30% for two thirds of global reaches and will be dominated by bias. Separate river discharge estimates will combine both SWOT and in situ data; these “gage constrained” discharge estimates can be expected to have lower systematic uncertainty. Temporal variations in river discharge timeseries will be dominated by random error and are expected to be estimated to within 15% for nearly all reaches, allowing accurate inference of event flow dynamics globally, including in ungaged basins. We believe this level of accuracy lays the groundwork for SWOT to enable breakthroughs in global hydrologic science.

This study intends to integrate heterogeneous remote sensing observations and hydrological modelling into a simple framework to monitor hydrological variables in the poorly gauged Congo River basin (CRB). It focuses on the possibility to retrieve effective channel depths and discharges all over the basin in near real time (NRT). First, this paper discusses the complexity of calibrating and validating a hydrologic–hydrodynamic model (namely the MGB model) in the CRB. Next, it provides a twofold methodology for inferring discharge at newly monitored virtual stations (VSs, crossings of a satellite ground track with a water body). It makes use of remotely sensed datasets together with in-situ data to constrain, calibrate and validate the model, and also to build a dataset of stage/discharge rating curves (RCs) at 709 VSs distributed all over the basin. The model was well calibrated at the four gages with recent data (Nash-Sutcliffe Efficiency, NSE> 0.77). The satisfactory quality of RCs basin-wide (mean NSE between simulated discharge and rated discharge at VSs, NSEmean = 0.67) is an indicator of the overall consistency of discharge simulations even in ungauged upstream sub-basins. This RC dataset provides an unprecedented possibility of NRT monitoring of CRB hydrological state from the current operational satellite altimetry constellation. The discharges estimated at newly monitored locations proved to be consistent with observations. They can be used to increase the temporal sampling of water surface elevation (WSE) monitoring from space with no need for new model runs. The RC located under the fast sampling orbit of the SWOT satellite, to be flown in 2022, will be used to infer daily discharge in major contributors and in the Cuvette Centrale, as soon as data is released.