AtmoScale helps meteorological, hydrological, and environmental institutions turn radar, satellite, and model data into cloud-native, decision-ready products — so your institution stays in control of its own data.
809 weather radars operate across 92 countries , generating some of Earth's most valuable environmental data.
Most of it remains locked in fragmented archives, vendor-specific formats, and workflows never designed for multi-year analysis or AI. Institutions end up adapting their science to whatever their storage will allow — instead of the other way around.
AtmoScale builds the cloud-native data infrastructure that gives that control back.
Radars send pulses; storms send echoes back. AtmoScale turns those returns into an analysis-ready, cloud-optimized cube — ready to feed any application, so people can act while it still matters.
Radar, satellite, and model signals from networks across your territory — flowing in continuously.
Our workflow harmonizes, chunks, and indexes them into cloud-native data cubes — analysis-ready.
The same cubes feed forecasts, QPE, alerts, research, and decision systems — all from one source of truth.
Benchmarks from the Radar DataTree platform — translated from raw speedups into what they mean for the people using the data.
Single-day quasi-vertical profile computation that took 8 hours on a conventional workflow now finishes in around 10 minutes. Same data. Same science.
A six-month quantitative precipitation estimation workflow that once required over a month of compute now completes in hours. Deep archives finally become tractable.
Analysis-ready, chunked formats mean the same computation transfers roughly a sixth of the data. Lower egress costs, faster workflows, smaller footprint.
Working prototypes on national radar archives from Colombia, the United States (NEXRAD), Germany, Italy, Canada, and Panama. Different vendors, different eras, one workflow.
Source: benchmarks from the Radar DataTree platform (Ladino et al., in prep.). Hardware and archive details available on request.
For too long, institutions have adapted their science to whatever their storage, vendor, or format would allow. That's backwards. The data exists to serve the mission — the mission shouldn't have to serve the data.
When your archive lives in a proprietary binary only one vendor can read, you have possession of the data — not governance. Open, self-describing, cloud-native formats put provenance, interoperability, and long-term control back with the institution.
Faster access, better formats, and cleaner pipelines are all in service of one thing: helping forecasters, scientists, and agencies make defensible decisions — in real time, or across decades of climate record.
Weather and water data often carry a public mandate. The infrastructure that delivers them should reflect that: open standards, open source, transparent methods, and no vendor lock-in by default.
— core conviction“The data is already valuable.
We make sure you're the one in control of it. ”
We don't just move bytes — we understand the observations behind them. Decades of published research in radar meteorology, hydrology, and atmospheric science, including the co-author of Radar Meteorology: A First Course .
Any vendor, any format, any source — radar, satellite, NWP, or surface observations. Built on open, cloud-native foundations (Zarr, Icechunk, xarray) so your institution never depends on us to read its own archive.
SaaS platforms, managed deployments, consulting engagements, hands-on training, or pilot projects — shaped around how your institution actually works, not how a product page wishes it did.
AtmoScale is a small team of atmospheric scientists, software engineers, and cloud-data specialists. Between us, decades of published research in radar meteorology and hydrology — and just as many years building the systems that actually move the data around.
We watched too many institutions invest in radar networks and then lose access to their own archives — stuck behind vendor formats, brittle pipelines, and infrastructure that wasn't built for the questions they actually want to ask.
Open formats. Cloud-native foundations. Hardware-agnostic by default. We build with you, not around you — pilot projects, managed deployments, training, or consulting, shaped to your institution's reality.
Modernizing a national radar archive. Building a cloud-native data platform. Preparing observations for AI. Wherever you are in the journey, let's talk.
Or just hit reply to any email from us — we read everything.