87th AMS Annual Meeting

Tuesday, 16 January 2007: 4:00 PM
Early Detection of Climate Change Signals in Local Historical High Frequency Precipitation Series
214B (Henry B. Gonzalez Convention Center)
Bernardo Gozzini, IBIMET - CNR, Sesto Fiorentino, Italy; and G. Maracchi, F. Meneguzzo, M. Pasqui, G. Menduni, B. Mazzanti, and F. Volpini
Poster PDF (2.4 MB)
The subject, methodologies and findings of the present research represent a step forward in the frame of a multi-year project supported and co-funded by the Arno River Basin Authority, Italy, the “Climate Reanalysis and Prediction over the Arno River Basin”, in cooperation with the Institute of Biometeorology, National Research Council, Italy, and the Foundation for Climate and Sustainability, Italy. The broad purposes of the research concern the quantitative analysis of the variability and tendencies of the extreme storms, as well as the droughts, both in the past and the future, and the previously used data bases consisted of historical series of daily precipitation records, as well as annual extremes at sub-daily durations. The present research represents a breakthrough due to the acquisition and analysis of historical time series of higher frequency data (hourly), which allows far more robust and significant statistical analyses, even when compared with the current literature and practices. Two long historical series of hourly precipitation data were digitalized starting from analogical records, one next to the coastline and the river mouth, the other in the inland mountains, covering 1945-2002 and 1930-2005, respectively. Data quality as well as homogeneity were checked by means of few statistical tests, such as the run test, including the comparison with data colleted at nearby rain gauges. The analysis of the overcomes of specific hourly thresholds, computed on the basis of the reference period 1961-1990, has shown a significant increase at both sites in the latest 15 years with regards to the previous times, as well as an increased seasonality, resulting in the concentration of most of the heavy rainfalls in September-November. A strong fall of the rainy hours has been observed at both gauges, mainly due to winter drying, and, limited to the mountainous gauge, a relevant increase of the average hourly precipitation intensity. The analysis of reference percentiles for the hourly precipitation, such as, 99°, 99,5° e 99,9°, has shown positive trends at both sites, as well as the precipitation series consisting of 3, 6, 12 and 24 hour events. In an attempt to provide some preliminary interpretation to the findings, the tendencies of heavy sub-daily rainfalls has been compared to the sea surface temperature (SST) time series, leading to partially significant outcomes and paving the way to further research. The findings are in the process of being extended to further time series, and are already contributing to the knowledge basis and even to the redefinition of the quantitative parameters of the hydraulic and hydro-geologic defense works (flash floods, floods, landslides and debris flows).

Supplementary URL: