We use ocean data and machine learning
to predict weather 3 weeks to 9 months ahead
more accurately than anyone else
What We Do
Weather Forecasts Worth Their Salt
At Salient, we use ocean data (such as sea surface salinity readings) and machine learning to predict precipitation and temperature from 3 weeks to 9 months ahead.
We predict heat and cold waves, floods and droughts far in advance so you can plan for what's coming. Backed by peer-reviewed research and engineered with cutting-edge machine learning, our technology enables better decision-making for a wide range of industries.
Customers use Salient's forecasts for diverse applications across agriculture, energy and utilities, insurance, finance and commodities, retail, natural disaster planning and beyond.
Raymond Schmitt, PhD
40-year career in physical oceanography research and technology
Scientist Emeritus, Woods Hole Oceanographic Institution
NASA Earth Science Advisory Committee
>100 refereed publications, >10k citations, h-index 48
Sam Levang, PhD
Co-Founder, Chief Scientist
Ph.D. Climate and Ocean Science, MIT/WHOI Joint Program
NASA Earth and Space Science Fellow (NESSF)
Expert in the global water cycle and its shifts with climate change
Why Salient Produces The Most Accurate Forecasts Ever
Our models focus on the ocean and land surface, which control seasonal weather patterns. The oceans drive the global water cycle, and they are rich in data that inform precipitation and temperature conditions over land.
Salient's machine learning platform avoids the bottlenecks of physics-based models, which must solve the chaotic atmosphere.
Drawing insights from historical conditions and future trends, from machine learning and physical models, our models produce the best long-range forecasts available. Our models have been independently validated in the US Bureau of Reclamation's "Subseasonal Forecast Rodeo", winning top prize for the most accurate rainfall forecasts.
North Atlantic salinity as a predictor of Sahel rainfall, Science Advances.
Centennial Changes of the Global Water Cycle in CMIP5 Models, Journal of Climate.