2026
Daily Harvest
-
Oct 4, 0 papers — No relevant papers from top-tier journals on this date.
-
Oct 3, 3 papers — New Earth-system model simulations show global temperature response to land-surface change scales linearly with planetary albedo shifts, while flash drought events are linked to amplified dust loading globally.
-
Oct 2, 1 papers — Deep learning of fossil pollen morphology reveals 25,000 years of grass diversity and precipitation changes in eastern African montane grasslands.
-
Sep 30, 3 papers — Antarctic Ice Sheet very likely committed to mass loss this century regardless of emissions scenario—a machine-learning framework now pins projections to satellite observations, constraining sea-level rise to up to 25.4 cm by 2100 (Nature Geoscience).
-
Sep 29, 5 papers — BridgeSAR uses Sentinel-1 SAR to measure water levels at bridges from space—opening new routes to continental-scale flood monitoring at thousands of ungauged sites.
-
Sep 28, 4 papers — Global observations from ~19,000 catchments show hydrological drought diverges sharply from atmospheric drying, with nonlinear catchment processes weakening or amplifying drought signals before they reach streamflow.
-
Sep 27, 1 papers — AeoLiS coastal dune model successfully applied to Grand Canyon sandbar, bridging coastal and fluvial aeolian morphodynamics for river management (Geophysical Research Letters).
-
Sep 25, 4 papers — CMIP6 models systematically overestimate global flood synchronization due to biased runoff peak timing, while a new River Wave framework links gauging-station flow data to river ecosystem dynamics.
-
Sep 24, 7 papers — SWOT satellite data reveals how humans have altered rivers worldwide while challenging flood prediction models, alongside new evidence that freshwater discharge from extreme precipitation drives open-ocean biogeochemical anomalies hundreds of kilometers offshore.
-
Sep 23, 1 papers — Paired OCO-3 and ECOSTRESS satellite observations capture diurnal ecosystem water-use efficiency, including midday drought stress, across vegetation types and climate zones.
Weekly Literature Review
-
Week 40 (Sep 28 - Oct 05), 3 papers, 3 papers — A watershed-agnostic explainable ML framework generalizes distributed hydrologic modeling across ungauged basins without basin-specific calibration.
-
Week 39 (Sep 21 - Sep 28), 1 paper, 1 papers — Deep learning data assimilation of sun-induced fluorescence into the ISBA land surface model enables more accurate vegetation-hydrology coupling in Earth system simulations (Biogeosciences).
-
Week 38 (Sep 14 - Sep 21), 2 papers, 2 papers — TerM LSM projects >50% surge in Limpopo peak flows under SSP5-8.5, while ORCHIDEE gains explicit Arctic shrub PFTs that cut simulated high-latitude biomass estimates by 13.5%.
-
Week 36 (Sep 01 - Sep 08), 5 papers, 5 papers — ORCHIDEE gains improved soil macroporosity and snow densification physics this week, alongside new LSM parameterizations for leaf turnover and rootzone soil moisture bias in hyper-resolution simulations.
-
Week 35 (Aug 24 - Aug 30), 0 papers, 0 papers — No relevant papers found this week.
-
Week 32 (Aug 03 - Aug 10), 0 papers, 0 papers — No relevant papers found this week.
-
Week 31 (Jul 27 - Aug 03), 2 papers, 2 papers — A new frozen-soil infiltration scheme in Canada’s operational SVS land surface model cuts severe streamflow simulation failures at cold-region stations from 33% to less than 1%; TIPMIP proposes a new CMIP7-candidate ESM protocol for probing Earth-system tipping points under controlled warming and negative-emission pathways.
-
Week 30 (Jul 21 - Jul 27), 19 papers, 19 papers — LSTM-based hybrid models outperform both process-based and purely data-driven approaches when extrapolating to warmer climates; benchmarking four reservoir-operation schemes for large-scale models reveals that added complexity only helps when auxiliary data are available.
-
Week 29 (Jul 13 - Jul 20), 2 papers, 2 papers — Both climate and landscape co-control streamflow behavior across European catchments, overturning the US-centric assumption that climate alone is the primary driver.
-
Week 27 (Jun 30 - Jul 7), 19 papers, 19 papers — An autonomous AI agent successfully closed the calibration–simulation loop for a hydrologic model without human intervention, signaling a new paradigm for automated model development.