I design early-warning and anticipatory-action systems for Bangladesh — translating meteorological forecasts into hazard-specific, location-specific triggers that funders, humanitarian actors, and communities can act on ahead of time.
I work on crisis anticipation and disaster risk financing with Start Network Bangladesh and FOREWARN Bangladesh, hosted by ESDO — contributing to programmes including Start Fund Bangladesh (SFB) and Disaster Risk Financing (DRF).
My work sits at the intersection of operational meteorology, climate data engineering, and decision-support design: multi-hazard threshold frameworks, ensemble forecast pipelines, cascading-impact models, and the dashboards and briefs that put them in front of the people who need to act.
Systems built end-to-end — from raw NWP/reanalysis data to the thresholds, dashboards, and briefs that decision-makers actually use.
A WMO-aligned, multi-hazard impact-based warning system covering 600 upazilas, blending four NWP ensembles across 13 variables into an operational risk dashboard.
Maps hydro-meteorological events through primary and secondary hazard chains to sector-level impacts and early actions — piloted for Chattogram District.
35 years of reanalysis climatology (1990–2025) for 600 upazilas, with derived heat-stress variables including WBGT, feels-like temperature, and relative humidity.
The operational dashboard itself — explore live multi-hazard risk across 600 upazilas, built on four blended NWP ensembles and a WMO 4×4 risk matrix.
Training material and walkthroughs on impact-based forecasting concepts and the technical pipelines behind them.
Warning → Impact-Based → Impact Forecasting paradigms, translated with precise Bengali/English terminology for upazila-level decision-making.
A tutorial series on building an hourly BMD SYNOP data pipeline with Google Apps Script, from decompression to structured sheets.
A hands-on walkthrough of pulling and processing ERA5-Land reanalysis data for hazard climatology at scale.
How four NWP ensembles, thirteen variables, and a WMO 4×4 risk matrix come together into one operational dashboard.