Weather data

VIPS uses weather data from various sources into models to calculate forecasts for plant pests. Some models calculate alerts automatically, while other models use user-selected weather data to provide personalized alerts or alerts based on form-based models.

Weather stations in LMT


The weather stations affiliated with the Agricultural Meteorological Service (LMT) provide meteorological data for all relevant weather parameters used in VIPS-related models. This is the traditional source of weather data used in VIPS. All automatic warnings found in the map on the front page are calculated with data from these weather stations.  

Most stations in the LMT network are well equipped with sensors that cover the specific needs of the various models.  

Private weather stations

Several of the models in VIPS can provide alerts based on data from private weather stations from several suppliers. The use of a private weather station is only possible if the station provides data for all weather parameters included in the model in question. Read more about the use of private weather stations here.

Some models, which are connected to VIPS via external services, cannot be run as private notification through VIPS. This applies to models from RIMpro (Apple Scab and Apple Wrapper), as well as grey mould in the Strawberry Advisory System and the Grôvfor model.

 

Point based data

VIPS users can choose to use weather data from a custom geographical point as a data basis for a personal forecast or in a form-based model. The point is linked to archived forecast data from the Norwegian Meteorological Institute (MET) via the OpenMeteo service. These data are hourly values with 1x1 km resolution. If a point outside Norway is chosen, weather data from other sources that may have a different spatial resolution are used.

 

Grid based data

Grid weather data is used in the risk maps. These are reanalysed data from MET, where forecasts have been adjusted with observations. The data is therefore somewhat more accurate than archived forecasts. For Norway, these data are provided as hourly values with 1x1 km resolution. Files with hourly data are retrieved from MET via an API, and built together into daily files with 24 hour values and stored in LMT's database.

 

Forecasts

Weather forecasts for use as input to the models are retrieved from MET and correspond to the forecasts found in yr.no.  

Short-term forecast delivers weather forecasts with hourly values 66 hours ahead. In practice, this means that we can calculate warnings for models that require hourly values two days ahead.   

Long-term forecasts (10-day forecasts) have a coarser temporal resolution (3, 6 and 12 hours), and can only be used to calculate forecasts for models that use daily values as a basis for calculation. This means that we can run calculation of alerts for models that use daily values as input data up to 9 full days in the future.  

The 21-day forecast uses the 10-day forecast for the first 10 days of the 21-day forecast and therefore corresponds to the long-term forecast for this period. After the first 10 days, datasets for weekly and daily values are used. The uncertainty is high, and this forecast is best suited for temperature-based models.  

Any interpolation of forecast data down to hourly values would give too much uncertainty for us to consider.

 

Normal values

Daily values based on the 30-year normal are available for places where there are weather stations. This data is used to calculate forecasts further ahead than weather forecasts can provide. E.g. when calculating flowering time in oats.

 

Calculation of missing values

Weather stations may at times lack data as a result of technical errors. Short-term gaps in the time series are replaced by re-analysed data for the site. This data is obtained from the MET.  

Weather forecasts provide data for a variety of weather parameters, but there may be models that require data that is only available from well-equipped monitoring stations. Some weather parameters can be calculated based on other data. This applies:  

Leaf moisture – calculated based on air temperature, humidity, precipitation and wind.

Evaporation   – calculated based on air temperature, humidity, radiation, wind and month.