FIspace

BO-MO is one of open-call partners in FIspace project, developing Weather Scenario App within Crop Protection Information Sharing trial.

Weather Scenario App

Weather Scenario App is a web service which delivers weather scenarios (combination of past weather and weather forecasts) for each requested location in the Netherlands in a high spatial resolution and a temporal resolution of one hour for up to one year ago and nine days ahead.

As already written weather scenario is done by combining weather forecasts with past weather data. Short range weather forecasts[1] are done by the Weather Research and Forecasting model (WRF), using Global Forecast System (GFS) as an input, while for medium-range forecasts GFS forecasts are used. Past weather is based on Royal Netherlands Meteorological institute (KNMI) stations data.

Weather Scenario calculation

Map of KNMI stations. (http://www.knmi.nl/klimatologie/images_algemeen/stations.jpg).

Map of KNMI stations.

For past weather 35 public available KNMI stations data are interpolated to rectangular grid using geostatistical methodology universal kriging with dependency on latitude, longitude, height and their squares and cross products. Output horizontal resolution is 0.05° (approximately 5.5 km). Data are updated daily and are available around 12 UTC of next day. Data are available from 1st of January 2014.

Available variables are: Temperature, Relative Humidity, Rainfall, Pressure, Specific humidity, wind speed at 10m and reference potential evapotranspiration following Penman-Monteith.

Short-range weather forecasts are done by dynamical downscaling (double nesting) of GFS forecasts using the WRF model, first on a larger domain with coarser resolution and secondon a smaller domain witha finer resolution as shown of Figure 3. The results are high resolution forecasts on approximately 6.2 km. Data are updated once per day and are available around 20 UTC for next 3 days.

WRF domains.

WRF domains.

For medium-range weather forecasts GFS data are interpolated to rectangular grid using geostatistical methodology universal kriging with dependency on latitude, longitude, height and their squares and cross products. Output horizontal resolution is 0.05° (approximately 5.5 km). Data are updated daily and are available around 17 UTC for next 9 days.

Available variables for weather forecasts are: Temperature, Relative Humidity, Rainfall, Pressure, Specific humidity, height of boundary layer, wind speed at 10m and reference potential evapotranspiration following Penman-Monteith.

Backend

The App backend provides its functionalities as a RESTful API, where request and response messages are in XML format. Request message contains information about time period (start and end date), location and requested weather variables. Based on this information backend prepares weather scenario data for nearest location as a combination of past weather data and short- and medium-range weather forecasts. Besides weather scenario response message contains information about time period, location and variables as well.

Location is following GML specifications and weather variables is following GRIB2 coding table.

Weather scenario backend provides two capabilities defined in FIspace core: PROVIDE_WEATHER_SCENARIO_SIMPLE and PROVIDE_WEATHER_SCENARIO (based on Data Reference Model Crop – drmCrop). The messages are defined on FIspace core component as well:

  • request message: WeatherScenarioSimpleRequest / WeatherScenarioRequest
  • response message: WeatherScenarioSimpleResponse / WeatherScenarioResponse

Widget, stand-alone webpage

In order to provide a graphical interface to allow farmer to interact with the App, simple widget for FIspace Wirecloud and stand-alone webpage has been prepared. They returns weather scenario in real-time for nearest point based on user input. User has to select:

  • start and end date
  • location on a map
  • variables

To access the standalone web-page user has to authenticate through SPT server of the FIspace platform.

Weather Scenario Stand-alone webpage.

Weather Scenario Stand-alone webpage.

Weather Scenario Stand-alone webpage: Result window showing results in Tabular view.

Weather Scenario Stand-alone webpage: Result window showing results in Tabular view.

Weather Scenario Stand-alone webpage: Result window showing results in Chart view.

Weather Scenario Stand-alone webpage: Result window showing results in Chart view.

[1]  More information about weather forecasts can be found in presentation given on SmartAgrimatics conference in June 2014 in Paris.