Automatic generation of a text forecast along a track

Belgingur Ltd. has created a novel weather forecasting framework, called Weather On Demand – WOD, that can be deployed in the cloud and customised for any location world-wide at a very short notice.

 

A recent addition to the WOD system is a routing forecast option that generates a simple text forecast along a track provided by the end-user.

 

The process is such that a user provides a list of coordinates, where each coordinate pair is accompanied by a timestamp, via an API.

Points of interest are identified along the track. Most commonly these points are the locations of weather stations, as they are generally placed where weather conditions are of interest and the WOD system has additional machine learning interpolation mechanisms in development for weather stations. From this set, along with on-the-hour locations, a representative, refined, lower resolution track is assembled, for which high-resolution forecast data is pulled.

From that forecast data, the information most relevant to the user is highlighted. Any difficult conditions, as well as a segmented summary is generated in simple, succinct text, programmable in any language.

An ongoing extension of this feature is to develop a module that can create a simple text forecast for any user defined region.

The WOD software is maintained in Git and can be installed on suitable hardware in a matter of hours, bringing the full flexibility and power of the WRF modelling system at your fingertips.