Tomorrow.io Announces $175M Financing to Deploy DeepSky, The World's First AI-Native Weather Satellite Constellation

X
Ruth Favela
By Ruth Favela
Ruth Favela
Ruth Favela
Ruth Favela is Tomorrow.io's AI Marketer. She draws on over 5 years of experience as an editor, writer, and social media manager for AI startups, B2B SaaS, and B2C products. In her role, Ruth focuses on using AI tools to create customer-first content for the various industries Tomorrow.io has solutions for. She writes about weather innovations, AI/ML modeling, weather API applications, weather AI use cases, and much more.
May 11, 2026· 5 min

How Tomorrow.io’s Microwave Sounders Are Improving Forecasting and Nowcasting

Ryan Honeyager on tropical cyclones, precipitation forecasting, and the growing operational impact of rapid-refresh microwave observations

As weather forecasting systems evolve, one thing is becoming increasingly clear: better forecasts depend on better observations.

In a recent webinar on Tomorrow.io’s Microwave Sounder Constellation, Ryan Honeyager, who leads Numerical Weather Prediction at Tomorrow.io, explored how the company’s Tomorrow Microwave Sounders (TMS) are being used in real forecasting applications — from tropical cyclone monitoring and precipitation nowcasting to numerical weather prediction and data assimilation.

His presentation highlighted how rapid-refresh microwave observations are helping fill critical gaps in the global observing system while opening new opportunities for operational forecasting agencies worldwide.

Why Frequent Microwave Observations Matter

Ryan began by recapping the core advantage of the TMS constellation: frequent global coverage.

The constellation consists of multiple small satellites operating across several low Earth orbits, collectively providing sub-hourly revisit rates around the world. That rapid refresh capability enables applications that historically were difficult with traditional microwave sounders, which often revisited locations only a few times per day.

Microwave sounders are particularly valuable because they can see inside clouds, unlike visible and infrared satellites that primarily observe cloud tops.

TMS observations allow forecasters to:

  • Infer vertical temperature profiles
  • Measure atmospheric water vapor
  • Analyze precipitation structures
  • Detect storm intensity signatures
  • Monitor atmospheric evolution beneath cloud layers

According to Ryan, this combination of cloud-penetrating capability and rapid revisit creates major advantages for forecasting and nowcasting applications.

Tracking Tropical Cyclones in Greater Detail

One of the clearest demonstrations of TMS capabilities came through tropical cyclone monitoring.

Ryan showcased overpasses of Typhoon Sinlaku, a Category 5 storm that impacted Guam. Using microwave composite imagery, TMS revealed:

  • Precipitation structures
  • Convective activity
  • Upper-level moisture fields
  • Storm-top proxies
  • Warm-core storm signatures

Because different microwave channels peak at different heights in the atmosphere, the system can build a layered picture of storm structure throughout the atmosphere — not just at cloud tops.

Ryan also highlighted Cyclone Laurel, which made four separate landfalls across Australia over a ten-day period. During that time, TMS captured 84 direct overpasses of the cyclone at multiple times throughout the day.

That level of temporal coverage is especially important for rapidly evolving storms, where observing changes between traditional satellite passes can be difficult.

Filling Gaps in the Global Observing System

A major reason TMS achieves this coverage is its orbital architecture.

Historically, most microwave sounders have operated in sun-synchronous orbits, meaning they pass over the same locations at roughly the same times each day. While effective for consistency, this creates temporal gaps in observations.

TMS supplements those traditional orbital planes with inclined orbits that precess over time, helping fill gaps in global coverage and creating a more evenly distributed observing network.

According to Ryan, this means there is almost always a recent microwave overpass available somewhere within evolving weather systems.

That becomes especially valuable for precipitation forecasting.

Improving Precipitation Nowcasting

One of the most important operational applications Ryan discussed was precipitation nowcasting.

Traditionally, precipitation retrievals have relied heavily on geostationary infrared observations. But infrared systems mainly infer precipitation indirectly from cloud-top temperatures.

Microwave observations can directly observe precipitation structures inside clouds.

Tomorrow.io has developed algorithms that fuse microwave observations with geostationary visible and infrared imagery to improve precipitation retrieval accuracy and forecasting skill.

According to Ryan, the “time value” of a TMS observation lasts roughly 70 minutes. That means precipitation forecasts initialized with microwave observations outperform geostationary-only retrievals for more than an hour after the observation occurs.

Because the constellation now delivers sub-hourly revisit globally, there are long periods where forecasting systems never need to fall back entirely on geostationary-only observations.

The result is improved precipitation nowcasting and stronger situational awareness during rapidly evolving weather events.

Supporting Tropical Cyclone Intensity Monitoring

Ryan also highlighted external operational applications already using TMS data.

Researchers at the University of Wisconsin are incorporating TMS observations into tropical cyclone intensity estimation systems alongside other microwave sounders. Their operational platform tracks cyclones globally, and TMS data already contributes roughly half of all observations within the system.

This frequent microwave coverage is particularly important for detecting rapid intensification events, which are becoming more common in a warming climate.

TMS and Numerical Weather Prediction

Beyond nowcasting, Ryan spent significant time discussing numerical weather prediction and data assimilation.

Because TMS is optimized for atmospheric temperature and water vapor retrievals, it is particularly well suited for assimilation into operational forecasting models.

Tomorrow.io developed a reference assimilation workflow using JEDI (Joint Effort for Data Assimilation Integration), a framework increasingly being adopted by agencies including NOAA and the UK Met Office.

The company’s assimilation studies explored:

  • Observation error characterization
  • Bias correction
  • Clear-sky and all-sky assimilation
  • Quality control approaches
  • Data thinning and averaging methods

Importantly, the workflows were designed to be understandable and accessible for operational scientists — not just software engineers.

Comparing TMS Against Operational Government Sounders

To evaluate performance, Tomorrow.io conducted data denial experiments comparing forecast impacts from TMS observations against NOAA’s Advanced Technology Microwave Sounder (ATMS), currently one of the primary operational government microwave sounders.

The results showed that while ATMS still outperforms TMS in some temperature-channel applications due to its larger instrument design and lower noise levels, TMS performs comparably in many water vapor applications.

Ryan explained that this reflects underlying instrument physics, including similar footprint sizes and channel bandwidths between the systems.

The studies also showed that TMS forecast impacts remain detectable out to approximately three days in forecast lead time.

Independent Evaluations and Growing Operational Adoption

Ryan concluded by highlighting several independent evaluations currently underway.

These include:

  • NOAA’s Commercial Microwave Data Pilot
  • JCSDA evaluation studies for the US Air Force
  • Ongoing NOAA and NASA assimilation studies
  • Additional evaluations related to the NASA TROPICS mission

NOAA has already integrated TMS observations into systems including:

  • Microwave Integrated Retrieval System (MiRS)
  • AWIPS workflows
  • Operational evaluation environments

According to Ryan, early results from these programs continue to demonstrate meaningful forecast improvements from rapid-refresh microwave observations.

Building the Future Global Observing System

Ryan’s presentation reinforced a broader shift taking place across weather forecasting.

As forecasting systems increasingly depend on higher-frequency global observations, rapid-refresh microwave constellations like TMS are becoming more operationally valuable — not only for research, but for real-world forecasting, nowcasting, and early warning systems.

By combining frequent global coverage, cloud-penetrating observations, and operational integration pathways, Tomorrow.io’s Microwave Sounder Constellation is helping build what may become a foundational layer of the next-generation global observing system.

See Tomorrow.io in Action!

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

TOS(Required)

Sign Up for the Tomorrow.io Weather API for Free

Try Now
Summarize this article with: Summarize with: