Improving life on Earth with data from space š
Joined May 2012
- Tweets 2,796
- Following 727
- Followers 14,377
- Likes 976
734 Photos and videos
May 29
Sub-seasonal forecasting, especially at weeks 3-6, is a notoriously difficult horizon for skill to be added. Physics-based models struggle to beat climatology in that window.
Most sub-seasonal forecasts, including the commercial ones, are post-processed ECMWF sub-seasonal range (S2S) forecasts. When the market is all reading the same signal, there's no edge.
Spire AI-S2S is 100% independent of any public sub-seasonal model and provides probabilistic forecasts out to 46 days, including weather regimes, anomalies, percentiles, and climatological exceedance probabilities. It's a 200-member ensemble built entirely in-house and designed specifically to maintain skill in the range where other models trend toward climatology.
ALT Spire AI-S2S weather regime forecasts for North America (top) and Europe (bottom) from our Cirrus Data Display Platform.
3
1
12
782
May 29
In a Jan. 1 - Feb. 15, 2026, verification against ERA5:
ā 14.76% improvement over ECMWF S2S* on 2-m temperature, weeks 3-6
ā 4.70% improvement over ECMWF S2S* on 500 hPa geopotential heights, weeks 3-6
ā Outperforms climatology by substantial margins across weeks 1-6
Both variables evaluated are critical for power and gas trading, as well as agriculture.
Recent improvements to initial conditions have led to meaningful gains, pushing results beyond what was previously published in a press release.
Read the full validation: bit.ly/4uDbwU2
ALT Across a 6-week verification period against ERA5, Spire's AI-S2S outperformed ECMWF'S S2S across all forecast lead times on both 2-m temperature and 500 hPa geopotential heights (pressure)*, the two fields key to forecasting energy demand and weather patterns.
1
2
9
610
May 29
Try it yourself. Over 3 months of reforecast history is now live on the @EarthmoverHQ Data Marketplace, and it's free to access, no subscription required: app.earthmover.io/marketplacā¦
*Spire conducted independent validation of data from its AI-S2S model, outside of the training and fine-tuning period, against ECMWF forecast data from January 1 - February 15, 2026. The ECMWF data used in this validation is published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license. These results are based on data and products of the European Centre for Medium-Range Weather Forecasts (ECMWF) ā Ā©2026 European Centre for Medium-Range Weather Forecasts (ECMWF). Source ecmwf.int.
1
7
522
May 27
We're excited to announce a strategic partnership with @SchaefflerGroup to build sovereign European space infrastructure ā industrialized in Germany, flight-proven in orbit, and deployable at scale for defense, weather, civil security, and critical infrastructure missions.
The partnership combines Schaeffler's precision engineering and manufacturing scale with our proven satellite platform expertise and extensive flight heritage, to establish a new European standard for space-industrial capability.
Read more: spire.com/press-release/schaā¦
3
4
29
1,110
May 21
Most energy trading desks run on the same public models, shrinking edge.
Today, we're announcing an expanded intraday-to-45-day forecast stack, purpose-built for energy trading desks and powered by our proprietary space-based data.
Validated against ERA5 reanalysis, Spire's AI-S2S ensemble model outperformed the European Center for Medium-Range Weather Forecasts (ECMWF) sub-seasonal-range (S2S) forecasts by 14.2% on surface temperatures at 3-6 weeks from January 1 to February 15, 2026.
1
3
15
1,444
May 21
"Weather is now the single biggest unpriced variable in energy markets ā and the traders who forecast it better win."
ā Shawn Mechelke, Spire GM of Weather, Climate, and Aviation
Stop trading on the same forecast as your competitors, see the weather differently, and act faster. Learn more: bit.ly/4usBlGu
#EnergyTrading #PowerTrading #PowerMarkets #WeatherIntelligence #RenewableEnergy
2
3
12
550
May 21
Adding the model validation chart for those who want a closer look at the numbers:
Over a 6-week verification period, Spire's AI-S2S showed a 14.2% improvement over ECMWF's S2S on 2-m temperature and a 3.23% improvement on 500-mb heights,* scored using CRPS (Continuous Ranked Probability Score; lower = higher forecast skill). Each model was scored on its native grid against ERA5 reanalysis.
*Spire conducted independent validation of data from its AI-S2S model, outside of the training and fine-tuning period, against ECMWF forecast data from January 1-February 15, 2026. The ECMWF data used in this validation is published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license. These results are based on data and products of the European Centre for Medium-Range Weather Forecasts (ECMWF) ā Ā©2026 European Centre for Medium-Range Weather Forecasts (ECMWF). Source ecmwf.int.
3
257
May 19
Airport operations run on real-time data ā and knowing where every aircraft is, at every moment, is foundational to everything else.
That's why we're excited to share we've been selected by @AmadeusITGroup to integrate our real-time ground and space-based aircraft tracking into the Amadeus Virtual Airport Operations Center (vAPOC), enriching the platform's operational intelligence with continuous global aircraft visibility to support daily operations and disruption management.
Read more: spire.com/press-release/spirā¦
1
2
10
619
May 14
One week out: weāre joining @EarthmoverHQ for a webinar you won't want to miss, and if you work in energy, insurance, or climate risk, that's especially true.
On May 21, Spire's Director of Weather Prediction and AI, Dr. Tom Gowan (@tommgowan), sits down with Earthmover CEO Dr. Ryan Abernathey (@rabernat) to break down our AI-driven sub-seasonal-to-seasonal forecast model (AI-S2S), which delivers forecasts and weather regimes up to six weeks ahead, updated daily, and built on data from one of the world's largest commercial satellite constellations.
In 45 minutes, we'll cover the science, the dataset, and real-world applications, which can be unlocked now through the Earthmover Data Marketplace.
š
Thursday, May 21 | ā° 11:00ā11:45 am ET | š Register here: m.earthmover.io/from-orbit-tā¦
ALT We're excited to announce a webinar collaboration with Earthmover titled "From Orbit to Earth: Unlocking Spire's AI-S2S Data at Scale." It is scheduled for Thursday, May 21, from 11:00 to 11:45 am ET, featuring Spire's Tom Gowan, PhD, and Earthmover's Dr. Ryan Abernathey.
1
1
10
428
May 7
We are excited to announce that our satellite manufacturing facility in Munich is now open and operational āĀ establishingĀ in-country satellite production to support sovereign space-based intelligence in Germany.
The facility features:
ā¢ An ISO-certified clean room
ā¢ End-to-end satellite manufacturing,Ā integrationĀ and testing capabilities
ā¢ Mission-specific payload development
ā¢ Capacity to produce up to 100 satellites per year
Read more: spire.com/press-release/spirā¦
4
18
696
Spire retweeted
Apr 30
šŖ Our Magquest deployable instrumentation boom, launched last month, has deployed on orbit š°ļø and is working as expected. Congrats to our customer @SpireGlobal for successfully meeting this important mission milestone!
#Space #Innovation #DeployableStructures #HugeinSpace
1
2
426
Apr 30
Brazil, Mexico, Chile: Mandatory climate reporting is here. Spire is the source, not the aggregator of weather data for the maritime industry. Meet us at Maritime Week Americas, May 18-21.
Meet Fitz-Robert Aaron, Sales Manager, Weather & Climate: meetings.hubspot.com/darija-ā¦
#MWA26 #MaritimeWeekAmericas #LATAM #Maritime
ALT Meet us at Maritime Week Americas, taking place from May 18-21 in Panama City. Fitz-Robert Aaron, Spire Sales Manager, will be representing our Weather & Climate team.
1
8
479
Apr 23
Weāre heading to Commodity Trading Week Europe in London (May 6-7)!
Most traders see weather as both a market risk and an opportunity, and desks are asking us how we can help them act on signals quickly.
Traders are using our space-based data to optimize strategies:
ā¢ Capture ramps others miss with High-Resolution and Optimized Point Forecasts
ā¢ Price risk weeks out with 45-day, market-moving weather signals with our AI sub-seasonal model
Our enhanced products show uncertainty clearly through model comparisons and trends, regime shifts, and Power Generation Forecasts in our one-stop, trade-ready intelligence dashboard, or by plugging APIs directly into your workflows.
If youāre looking to tighten risk and improve positioning, let's talk. Book a meeting: lnkd.in/gSnGngDs
#CTW2026 #CommodityTradingWeek
ALT Spire Weather & Climate experts attending Commodity Trading Week Europe, May 6ā7, London: Dennis Quick (Earth Insights Sales Manager), Gerald van der Grijn (Senior Product Manager), and Ram Vadlamani (Global Weather Solutions).
2
11
670
Apr 21
Last week, Super Typhoon Sinlaku struck US territories in the West Pacific as the equivalent of a Category 4 storm, making landfall across the Northern Mariana Islands with devastating force.
Impacts were severe:
ā¢ 140 mph wind gusts at landfall
ā¢ 10-15 ft storm surge and widespread flooding
ā¢ Power outages, infrastructure damage, and downed lines
ā¢ Prolonged impacts due to slow movement (~5 mph)
ALT Satellite image of Super Typhoon Sinlaku over the western Pacific as it approached the Northern Mariana Islands on April 13, 2026. At maximum intensity, the storm packed winds of 175 mph (280 kph), making it the equivalent of a Category 5 storm on the Saffir-Simpson Hurricane Wind Scale. Image credit: MODIS Land Rapid Response Team, NASA GSFC
3
2
15
847
Apr 21
At a 24-hour lead time, the picture sharpened further:
Global models (IFS, AI/ML, and ensemble guidance) skewed too far east and north, placing populated areas on the storm's weaker left flank and significantly underestimating impacts.
Spire's track was within 10 miles (16 km) of the actual landfall location on Tinian Island (30-50 miles or 50-80 km closer than other global models), correctly placing the most destructive eyewall impacts.
ALT Track Accuracy: At 24-hour lead time, Spire's forecast identified the correct westward track, outperforming leading global models. While others skewed east/north, Spire aligned with the observed path, improving confidence in landfall timing, location, and downstream impact planning. Image via Google DeepMind / CC BY 4.0 ā [source: https://deepmind.google.com/science/weatherlab], adapted from original
1
9
475
Apr 21
In high-stakes scenarios like Sinlaku, that margin matters:
ā¢ Infrastructure and energy grid protection depend on knowing where the worst winds will be
ā¢ Emergency response can be staged and mobilized ahead of the most destructive corridor
ā¢ An accurate forecast means protecting lives, reducing costs, and building resilience
#WeatherIntelligence #TyphoonSinlaku #SpireGlobal
2
298
Apr 17
Agriculture, one of the most climate-exposed industries, is still operating with critical environmental data gaps.
That shows up as:
ā¢ Missed early crop stress signals
ā¢ Inefficient irrigation
ā¢ Reactive decision-making
Spire is working to change that š§µā
1
2
18
4,587
Apr 17
āā¦strengthened our ability to deliver data-driven insightsā¦ enabling more precise, scalable, and timely decisions.ā
This is what closing the data gap actually looks like in practice.
1
3
418
Apr 17
This isnāt just better data.
Itās a shift to:
ā”ļø Earlier signals
ā”ļø Smarter inputs
ā”ļø More confident decisions
Turn soil moisture insights and advanced weather forecasting into a strategic advantage.
Request a live demo: spire.com/weather-climate/agā¦ā¦
#AgTech #PrecisionAgriculture #ClimateRisk #EarthObservation #SoilMoisture
1
5
343