Space–Time Characterization of Rainfall Field in Tuscany

• Main Author: Alessandro Mazza
• Format: Article
• Language: English
• Published: MDPI AG 2017-01-01
• Series: Water
• Subjects: precipitation distribution; extreme events; seasonality; rain gauge; kriging; stationary random function; exponential variogram model; trend detection
• Online Access: http://www.mdpi.com/2073-4441/9/2/86

Precipitation during the period 2001–2016 over the northern and central part of Tuscany was studied in order to characterize the rainfall regime. The dataset consisted of hourly cumulative rainfall series recorded by a network of 801 rain gauges. The territory was divided into 30 × 30 km2 square areas where the annual and seasonal Average Cumulative Rainfall (ACR) and its uncertainty were estimated using the Non-Parametric Ordinary Block Kriging (NPOBK) technique. The choice of area size was a compromise that allows a satisfactory spatial resolution and an acceptable uncertainty of ACR estimates. The daily ACR was estimated using a less computationally expensive technique, averaging the cumulative rainfall measurements in the area. The trend analysis of annual and seasonal ACR time series was performed by means of the Mann–Kendall test. Four climatic zones were identified: the north-western was the rainiest, followed by the north-eastern, northcentral and south-central. An overall increase in precipitation was identified, more intense in the north-west, and determined mostly by the increase in winter precipitation. On the entire territory, the number of rainy days, mean precipitation intensity and sum of daily ACR in four intensity groups were evaluated at annual and seasonal scale. The main result was a magnitude of the ACR trend evaluated as 35 mm/year, due mainly to an increase in light and extreme precipitations. This result is in contrast with the decreasing rainfall detected in the past decades.