Hyper-Local Weather
Weather affects most aspects of our lives, yet the importance of accurate, hyper-local surface and atmospheric weather forecasting is often overlooked.
Transforming prediction
Surface-level forecasts help drivers navigate more safely, fleets optimize routes, cities maintain roads, event planners improve safety, and farmers grow crops that flourish.
Weather forecasts are more than predicting the conditions of the atmosphere above us, they’re also about understanding the true conditions where we all reside: at the ground level.
Surface Weather Forecasts
Typical weather forecasts do not offer accurate information about how precipitation is going to affect pavement surfaces, roads and soil.
In instances where this information is provided, many of those forecasts are simply based on mapping a single atmospheric precipitation forecast to a location and inferring the impact to the ground or road surface. In other words, precipitation type and rate converts to ground condition state. This surface forecast approach is too simplistic and produces false positives, leading to costly delays, accidents and maintenance issues.
The Fathym Forecaster relies on cutting-edge machine learning from a global network of fixed and mobile sensor observations, and a proprietary method of blending sensor data with multiple model output datasets to create highly accurate, hyper-local pavement and atmospheric forecasts.
Scalable Forecasting
Fathym’s statistically-based, blended model allows forecasts to be dynamically generated at high speeds, meaning their real-time relevancy and accuracy is vastly improved from traditional physical forecast models.
Statistical modeling within the Fathym Forecaster relies on a smaller dataset for model training, which enables the model to generate more consistent results.
Physical models are data driven. They require more raw input for training on a specific forecast point (road type, road base, heat transfer, shade, UV), which yields slower processing cycles and limits scale across cloud resources. Although the physical model will provide a more accurate forecast at one specific point, it loses its value as the user scales across thousands or millions of points along a road segment. The statistical model consistently outperforms the physical model in this use case.
Fathym’s road temperature forecast has been proven to be within 0.5°C during the morning and overnight hours and within 1°C during the heat of the day out to 52 hours, which exceeds the accuracy of physically-based road weather models deployed today.
Dynamically-Generated Points & Routes Forecast
Forecasts are available as single geo-locations (such as a farm or stadium), a series of points (such as a city) and routes. The forecast is available through a simple API.
Take Your Forecast Further
Harness Fathym’s low-code IoT framework to combine surface weather forecasting with data visualisation widgets, mapping and routing services and data triggered alerts.
