Severe hydro-meteorological events are increasing in frequency and magnitude. The societal and economic implications of these events, including loss of life and property damage, are the prime motivation for DRIHM’s commitment to contribute towards the following aspects:
- Supporting Civil Protection decision makers with reliable information about where extreme meteorological events are most likely to occur;
- Protecting people and infrastructure from the direct impact of severe weather;
- Targeting operational rescue activity at areas known to be at high risk.
In this context, short and medium term forecasting and management of severe hydro-meteorological events are highly topical and represent an important contribution to the procedures implemented by civil protection authorities. The peculiar nature of severe hydro-meteorological processes occurring in most small and medium-size catchments in complex orography areas, such as the Mediterranean region, make them difficult to predict. This is due to a number of reasons:
- The complex orography produces event response times in the order of a few hours;
- Flash floods develop rapidly during the rainy season and suddenly inundate the terminal flood plains, incorporating many historic cities;
- Traditional warning systems based on rainfall observations and rainfall–runoff modeling cannot provide the timely predictions needed to implement the required precautionary civil protection measures;
- Social safety demands that hydrologists reliably predict ground effects 24 hours in advance and this requires using rainfall predictions as input to rainfall–runoff models, resulting in the use of a full hydro-meteorological forecasting chain;
- Many sources of uncertainty are associated with hydro-meteorological forecasting chains.
The predictive ability of these severe hydro-meteorological events can be improved with:
- Coherent and comparable observations at multiple locations;
- Denser observations in time and space;
- Better access to data;
- More detailed models;
- Combination of different modeling tools and post-processing tools;
A ‘joined-up science’ multidisciplinary perspective. Therefore, a key challenge for current HMR is to develop and validate new tools and methods to meet these needs. It is an interdisciplinary endeavor, requiring collaboration between measurement scientists and agencies collecting data, meteorological scientists forecasting the weather, and hydrological and hydraulic scientists predicting runoff and flooding and public services such as national environmental and civil protection agencies. It is also an international endeavor, requiring an integrated view of events that freely cross international borders. In the recent years, the quantity and complexity of the modeling tools and datasets have increased dramatically:
- The availability and quality of remote sensing observational data from satellites and ground-based radars (providing complete three-dimensional coverage of the atmospheric and land surface state) has vastly increased;
- The growing use of ensemble forecasting methods that combine multiple numerical weather prediction and hydrological models to quantify the uncertainty in the forecast, has dramatically multiplied the computational costs;
- There is increasing recognition of the need to understand the entire flood forecasting chain, from observations through to civil protection response. This calls for the deployment of complex workflows able to combine different data sets, hydro-meteorological models and local decision-making expertise in a flexible manner.
However, as suggested also by the DRIHMS project:
- Progress in tackling this HMR challenge is slowed by the difficulty in accessing data that are scattered in different archives, different countries, and different formats;
- Collaboration is inhibited by the need for complex weather and hydrological models with significant high performance computing requirements.
It is understandable that lack of data limits the researchers’ ability to test and further advance models, resulting in a compromise that restricts their attention to the data and models locally available. However, the benefits of comparing and combining the full range of existing datasets and tools are clear.
With these goals in mind, the DRIMH project starts.