Work Metoc Forecasting Service

A visualisation tool to rapidly build forecasting products based on atmospheric, sea state and ocean weather data.

METOC Forecasting Service

Dynamic forecast maps, based on close-to-real-time data, continuously updated.

Forecasting

A meteorology agency needed a visualisation tool to rapidly build forecasting mapping products based on atmospheric, wave and ocean weather data published by the Australian Bureau of Meteorology (BoM).

An existing system was already in place but was available only to a closed group and process-intensive to generate products.

The agency wanted the system to consume and build dynamic forecast maps, based on close-to-real-time data, continuously updated from BoM.

Discover and define

The project involved R&D activities to discover and define solutions to some of the core underlying problems outside the capability of the existing system.

We wanted to support the agency in building a system that pushed new boundaries on current solutions. We proposed an R&D approach to design creative and intelligent solutions that were proven by prototypes.

BoM data is provided in a scientific data format called NetCDF (Network Common Data Form) via an open source server application called THREDDS. THREDDS serves NetCDF data as OGC services (WFS, WMS, WCS and other protocols).

Because the agency was already using ArcGIS Server products, we proposed utilising the existing ArcGIS infrastructure to consume THREDDS OGC services in building the forecasting mapping layers the agency required.

Incompatibilities between ArcGIS tools and THREDDS services meant we had to perform R&D and come up with new solutions to implement the forecasting tools generating contours, classified.

A simple interface

We wanted to build a core stateless service layer and a consumer web application with a simple interface. It also had to generate and visualise forecasting mapping products in a timely fashion.

To do this, we proposed building core capabilities, forecasting components, and tools as stateless web services that would be consumed by a simple web application.

The stateless service layer would support scalability and the web application would provide accessibility and usage by a wider audience with a capability to authorise features in the client intranet.

We built a single page javascript web application based on a MVP (Model View Presenter) framework called Backbone JS. We used modern Javascript libraries and techniques to build the application (practices defined in Common JS specifications)

Result

The new solution is currently being showcased and experimented around the agency. Feedback from various stakeholders is driving improvements towards production use.

Technologies