GarrCast produces gauge-adjusted radar rainfall — NOAA NEXRAD bias-corrected against your rain gauges, quality-controlled by hydro-meteorologists, and delivered to your basins with documented accuracy. The detail of radar, anchored to the truth of the gauge.
NEXRAD Level II · ~5-min updates · gauge bias correction · QC & verification · model-ready
A rain gauge measures rainfall accurately — but only over a few inches of ground, and it tells you nothing about what fell between gauges. Radar maps the whole storm, but estimates amounts indirectly and drifts without correction. GARR combines them: radar for spatial structure, gauges for ground truth.
→ GARR = radar's coverage, corrected to the gauges' truth, and quality-controlled.
Every event moves through the same disciplined pipeline — automated in real time, then reviewed and documented for the record.
NEXRAD Level II reflectivity (dBZ) and rain-gauge time series stream in continuously.
Reflectivity converted to rain rate via adaptive Z–R and dual-pol estimators.
Radar is adjusted against gauges to remove systematic bias and anchor amounts.
Anomalies removed; hydro-met review; accuracy documented vs. independent gauges.
Filtered to your grids, catchments & basins — model-ready, with reports.
Reflectivity-to-rain-rate relationships are continuously evaluated against gauges rather than fixed, capturing how storm type changes the conversion.
Dual-pol Level II variables improve rain-rate estimation and help screen hail and non-meteorological echoes.
Spatially and temporally varied bias correction ties the radar field to ground truth — even when some gauges drop out.
Automated screening plus manual hydro-meteorologist review removes inconsistent radar and gauge data per storm period.
GARR is cross-checked against independent gauges and references like NWS Stage IV, with documented accuracy.
Final rainfall is weighted and averaged to your catchments and basins — the geometry your models actually use.
Machine learning improves the hard, fuzzy parts of the problem — without ever replacing the physics-based, gauge-anchored, documented core that makes GARR defensible.
Data-driven relationships adapt to regional storm regimes and seasons for better rain rates.
Models capture nonlinear radar–gauge relationships across range and terrain.
Automatic flagging of clogged gauges, ground clutter, hail spikes, and artifacts.
Deep-learning extrapolation projects the QPE field forward for short-range forecasting.
Spatially variable rainfall for collection-system planning, design, rehabilitation, and operations — capturing how sewersheds respond to real storms.
Quality-controlled, documented inputs and daily wet/dry catchment assessments suitable for regulatory submission.
Historical GARR for recurrence-interval analysis, IDF context, and design-storm characterization.
Defensible, documented rainfall reconstruction for claims, disputes, and post-event "what actually fell here" questions.
High-resolution real-time rainfall as input to monitoring, alerting, and emergency response.
Watershed rainfall for inflow forecasting, gate operations, and water-resources planning.
Sewer and stormwater modeling, capacity analysis, RTC, and regulatory rainfall of record.
Stormwater and drainage design, flood awareness, and defensible rainfall documentation.
Calibrated, model-ready rainfall for H&H studies, master plans, and forensic analysis.
Basin rainfall for early warning, reservoir inflow, and operations.
Real-time and historical rainfall for response, corridors, and after-action review.
Independent, documented storm reconstruction for claims and disputes.
Send us your service area and the storms that matter — we'll walk you through GARR for your basins.
hello@garrcast.com