The Finnish Meteorological Institute (FMI) has launched a new pilot project in partnership with Finland’s National Emergency Supply Agency (NESA) to develop more accurate impact forecasts to guide emergency response and supply services.
The SILVA project uses a comprehensive database of two decades of historical information on weather and climate events and their subsequent impact to develop new forecast products that leverage advanced methods such as machine learning.
Ari-Juhani Punkka, project lead from the FMI, said, “The database contains weather-related events affecting sectors such as energy, transport and insurance. Many of the records on events cover many years, and are national in scope, so the data that has been collected is rare in its coverage even by global comparison.”
The new products will help those involved in activities that are critical in preparing for dangerous weather situations and to help plan for the right allocation of resources. It is hoped that the new products will make it possible to anticipate the number of emergency calls related to situations such as wildfires a few days before the event begins.
The impact forecasts are presented as time series or map images in which the significance of the weather events and its effects are indicated by a green to red color scale.
Tapio Tourula, emergency preparedness manager, NESA, said, “We are testing and applying a completely new approach to preparing and developing security of supply. Information from the memories of specific individuals and entities becomes shared digital data. A method based on modern artificial intelligence can produce new kinds of information and services and improve the continuity of critical activities in society.”
New products developed as part of SILVA are being piloted up until autumn of this year under real weather situations. The testing of the pilot products involves around 40 organizations in emergency services, energy, and various forms of transport.
“Our present aim is to collect feedback from the use and development of the new products as extensively as possible,” added Punkka. “During the pilot stage, there are also plans to hold briefings on how to use the new products when significant weather situations are about to occur, so there will be repeated interaction with the testers.”
There will also be efforts to use an open data interface to make historical impact data available to anyone who wishes to use it.
