Q & A

Answer:

Our systems run in our cloud platform on a continuous operation. The updated results are shown in a dashboard.
Answer:
We use weather forecasts, so our systems can reliably predict floods up to seven days ahead.
Answer:
The accuracy of the forecasts depends on the specific case (since it can be affected by data availability and quality). Nevertheless, as a rule of thumb our models have shown accuracies above 90-95% for historical data. Since the weather forecasts are not deterministic, our predictions will also have a prediction interval.
Answer:
Yes, this is exactly our main purpose and what distinguishes our systems from other forecasting systems (like Varsom). Our systems predict the parameters (ex: water level) locally. That information is presented in the dashboard along with the corresponding warning level. If the client choses to have 2D or 3D interactive maps, then even the water level on the lateral walls of buildings or infrastructures can be seen.
Answer:
Yes, on the 2D interactive map you can see the water depth in a given place by clicking in that position. On the 3D interactive map, you can see the water level on the lateral walls of buildings or infrastructures.
Answer:
Yes, an automatic alert is possible to be done either by SMS, email or though the mobile app.
Answer:
Yes, our systems can work in regulated catchments, as is the case of Sauda Flood Warning System (see Case Studies). To have good results we need to collaborate with the operating hydropower companies and have access to their data. In fact, our systems can help optimizing the hydropower production, by reducing the amount of water that needs to pass the spillway during floods.
Answer:
Yes, our systems can include snowmelt, as is the case of Sauda Flood Warning System (see Case Studies). In that case we were able to predict accurately the inflow to the hydropower reservoirs due to snowmelt.
Answer:
Our systems use historical data, but they do not rely on attributing any sort of frequency (return period). Therefore, they predict what is supposed to occur for a given event coming from weather forecast. This has been proven very reliable, for example in Tovdal Flood Warning System (see Case Studies) our system would have been able to predict the Oct. 2017 extreme flood event (highest ever in that catchment) with 4 days-ahead.