ALBERT

Description: Calibration and validation of flood risk maps at a national or a supra-national level remains a problematic aspect due to the limited information available to carry out these tasks. However, this validation is essential to define the representativeness of the results and for end users to gain confidence in them. In recent years, the use of information derived from social networks is becoming generalized in the field of natural risks as a means of validating results. However, the use of data from social networks also has its drawbacks, such as the biases associated with age and gender and their spatial distribution. The use of information associated with phone calls to Emergency Services (112) can resolve these deficiencies, although other problems are still latent. For example, a bias does exist in the relationship between the size of the population and the number of calls to the Emergency Services. This last aspect determines that global regression models have not been effective in simulating the behavior of related variables (calls to Emergency Services–Potential Flood Risk). Faced with this situation, the use of local regression models (such as locally estimated scatterplot smoothing (LOESS)) showed satisfactory results in the calibration of potential flood risk levels in the Autonomous Community of Castilla-La Mancha (Spain). This provides a new methodological path to the calibration studies of flood risk cartographies at national and supra-national levels. The results obtained through LOESS local regression models allowed us to establish the correct relationship between categorized potential risk levels and the inferred potential risk. They also permitted us to define the cases in which said levels differed ostensibly and where potential risk due to floods assigned to those municipalities led to a lower level of confidence. Therefore, based on the number of calls to the Emergency Service, we can categorize those municipalities that should be the subject of a more detailed study and those whose classification should be revised in future updates.