GWC’s technology provides sophisticated post-processing of existing numerical forecasts, improving on them through several algorithms.
GWC’s technology employs sophisticated algorithms to post-process numerical weather prediction models. These are the models (such as GFS and ECMWF) computed by national weather services. GWC’s automated algorithms apply machine learning to fine-tune the accuracy. This replicates the role of the human meteorologist, aggregating all available models and data to present a more accurate forecast than is possible from any individual model.
The core technology starts by making improved forecasts at sensor locations. A significant strength of our approach is the ability to ingest data from customer-owned sensors and use that data to enhance the forecast at those locations. In real-world tests, our technology outperforms competing forecast
The technology includes three major steps:
- a regression analysis to bias-adjust each model at the forecast location,
- a weighting of each model based on how well it is expected to perform under current conditions followed by combining the weighted model outputs, and
- a forward error correction that smoothly modifies the forecast over the next few hours to be consistent with current observation
Recently, this engine has been extended from sensor-based forecasts to one capable of providing forecasts globally at non-sensor locations – in other words, at any chosen latitude and longitude. The technology is updated on an ongoing basis through GWC’s R&D investments, including collaborative research with NCAR
GWC products are delivered through web services, and we provide a Software Development Kit (SDK) with a complete API set.
Our cloud-based architecture and SDK make it easy to customize our products for particular applications and customers. GWC provides consulting services to support these customer-specific enhancements/modifications to our standard forecast services.
- Custom websites with text- or map-based precision weather
- Website widgets that integrate weather into website functions
- Custom mobile applications
- Integration of region-specific or specialized numerical forecast models
- Integration of data from new and/or proprietary sensor networks
- Custom solutions to vertical- or customer-specific needs
GWC’s Weather SDK
For developers, GWC’s Weather SDK (Software Development Kit) is designed to easily integrate complex weather functions into custom applications. Functions and options can include:
Alerts – Be it a mobile, desktop or web application, our Weather SDK offers developers a variety of alerts and notifications.
Along-route weather – GWC’s weather forecast API allows developers to include complete along-route weather in any application platform.
Visualization/precision mapping – Our SDK creates and delivers weather data visualization in context with maps and other media.
Integration for web and mobile applications – Use our Weather API to get quick, easy, seamless forecast integration into any app on any device.
What is the accuracy of the GWC forecast?
The forecast accuracy has been tested repeatedly over the last decade and routinely proves to be the best available forecast across a variety of comparison approaches. Specific accuracy results depend on a wide variety of considerations, including location, season, and variable being assessed.
What forecasted variables are available?
See Specifications of Services in the document for each individual ConnectedWX service.
What underlying numerical weather models are included?
Relevant models are included from the U.S. National Weather Service (NWS), the European Centre for Medium-Range Weather Forecasts (ECMWF), the UK Met Office, and Environment Canada. The particular models used depend on the region and the forecast time.
Do some regions have higher fidelity?
Yes. We currently use more model data (for SensorWX) and perform intermediate computations at higher spatial resolution (for PointWX) in North America and Europe. Additional high fidelity regions can be developed on request.
What are the data delivery options?
SensorWX and PointWX are delivered as a web service using standard web protocols. They can be delivered as a pre-specified dataset, such as for a set of 10,000 cities delivered each hour, or as an individual forecast on a transaction basis triggered by a latitude/longitude request at a specified time.
What are the data file formats?
When accessed as datasets, SensorWX and PointWX data are currently delivered in CSV format through secure FTP. Delivery in other formats (such as XML and netCDF) can be accommodated on request. When accessed through our API, SensorWX and PointWX data are delivered in XML format through restful web services using the URI/http method.
What forecast horizons are available?
Hourly resolution for SensorWX and PointWX is currently available to 15 days (360 hours from forecast time). Shorter periods can be made available on request. Sub-hourly (5 or 15 minute) resolution is available for hours 1-3.
What is the data update frequency and latency?
Hourly forecasts are updated each hour. The hourly update incorporates the most recent underlying numerical model data and uses current forecasts to perform a forward error correction that ensures the forecasts do not deviate significantly from the current observations, as happens with other forecast methods. Data are available within 30 minutes of the forecast time. Sub-hourly forecasts are updated at 5 or 15 minute intervals.
Is it possible to create customized forecasts using customer-provided data or forecast engines?
Yes. A significant advantage of the modeling technology used by GWC is the ease of including customer-provided observation data or forecast engines, both of which can be implemented to enhance the quality of GWC forecasts.
Does GWC provide mobile or web clients to display the forecasts?
Yes. We have developed several web and mobile clients that we make available free to customers who use our forecast services. These are provided as part of our SDK and can be readily modified to support customer-specific needs. A variety of displays are available, including time series of all variables and time-sequence imagery of selected variables (radar, temperature, precipitation, and wind speed). Examples of both are shown here.
Does GWC provide imagery?
Yes. We provide both static (single image) and dynamic (image sequence over time) imagery. Current offerings include Doppler radar (U.S. only), temperature, probability of precipitation, and precipitation type. These are available through our standard API.
How reliable are the data services?
They are very reliable. Services are provided from geographically redundant data centers. Reliability has exceeded 99.9% over GWC operating history. SLA terms are available on request.