From Deterministic to Probabilistic Forecasts: The ‘Shift-Target’ Approach in the Milan Urban Area (Northern Italy)

Gabriele Lombardi,Marco Mancini,Silvio Davolio,Giovanni Ravazzani,Alessandro Ceppi

doi:10.3390/geosciences8050181

Abstract

The number of natural catastrophes that affect people worldwide is increasing; among these, the hydro-meteorological events represent the worst scenario due to the thousands of deaths and huge damages to private and state ownership they can cause. To prevent this, besides various structural measures, many non-structural solutions, such as the implementation of flood warning systems, have been proposed in recent years. In this study, we suggest a low computational cost method to produce a probabilistic flood prediction system using a single forecast precipitation scenario perturbed via a spatial shift. In fact, it is well-known that accurate forecasts of heavy precipitation, especially associated with deep moist convection, are challenging due to uncertainties arising from the numerical weather prediction (NWP), and high sensitivity to misrepresentation of the initial atmospheric state. Inaccuracies in precipitation forecasts are partially due to spatial misplacing. To produce hydro-meteorological simulations and forecasts, we use a flood forecasting system which comprises the physically-based rainfall-runoff hydrological model FEST-WB developed by the Politecnico di Milano, and the MOLOCH meteorological model provided by the Institute of Atmospheric Sciences and Climate (CNR-ISAC). The areas of study are the hydrological basins of the rivers Seveso, Olona, and Lambro located in the northern part of Milan city (northern Italy) where this system works every day in real-time. In this paper, we show the performance of reforecasts carried out between the years 2012 and 2015: in particular, we explore the ‘Shift-Target’ (ST) approach in order to obtain 40 ensemble members, which we assume equally likely, derived from the available deterministic precipitation forecast. Performances are shown through statistical indexes based on exceeding the threshold for different gauge stations over the three hydrological basins. Results highlight how the Shift-Target approach complements the deterministic MOLOCH-based flood forecast for warning purposes.

Highlights

From a civil protection point of view, hydro-meteorological forecasts can be seen as a powerful tool of non-structural measures to produce early flood warnings and better counteract potential river flood impacts, whose number is increasing worldwide [1]
In 1993, good forecasting is a matter of “getting it right”, and to make the receivers understand it, and, above all, to be able to draw conclusions from it [5]. In this analysis, we are not interested in predicting river discharge with an accurate flood peak in magnitude as well as timing, but in predicting the probability of exceeding any threshold before the event, in order to provide early flood warnings to local authorities. It is well known in the scientific community around the world that ensemble or probabilistic forecasts contain more information than single-valued forecasts [6,7], a key topic of the EFAS (European Flood Awareness System) and HEPEX projects “to demonstrate the added value of hydrological ensemble predictions (HEPS) for emergency management and water resources sectors to make decisions that have important consequences for economy, public health and safety [8]”
How do these two approaches exactly predict a non-event? In Table 4, we report the observed frequency related to the MOLOCH green code prediction, Correct Negatives Ratio (CNR), and the one related to a 95–100% predicted shift probability of the same green code

Summary

Introduction

From a civil protection point of view, hydro-meteorological forecasts can be seen as a powerful tool of non-structural measures to produce early flood warnings and better counteract potential river flood impacts, whose number is increasing worldwide [1]. In 1993, good forecasting is a matter of “getting it right”, and to make the receivers understand it, and, above all, to be able to draw conclusions from it [5] Adopting this framework, in this analysis, we are not interested in predicting river discharge with an accurate flood peak in magnitude as well as timing, but in predicting the probability of exceeding any threshold before the event, in order to provide early flood warnings to local authorities. It is well known in the scientific community around the world that ensemble or probabilistic forecasts contain more information than single-valued forecasts [6,7], a key topic of the EFAS (European Flood Awareness System) and HEPEX (https://hepex.irstea.fr/) projects “to demonstrate the added value of hydrological ensemble predictions (HEPS) for emergency management and water resources sectors to make decisions that have important consequences for economy, public health and safety [8]”

Objectives

Methods

Results

Conclusion