TITLE:
Liquid-Solid Partitioning of Precipitation along an Altitude Gradient and Its Statistical Properties: An Italian Case Study
AUTHORS:
Francesco Avanzi, Carlo De Michele, Antonio Ghezzi
KEYWORDS:
Climate Change; Statistical Inference; Precipitation; Events Partitioning; Snow; Mountainous Hydrology
JOURNAL NAME:
American Journal of Climate Change,
Vol.3 No.1,
March
21,
2014
ABSTRACT: Climate
change is a living topic when dealing with modern natural sciences. The
increase in the average air temperature, as measured in the last decades, is
considered as the most relevant effect of climate change on the Earth system.
Since the air temperature has a key role in determining the partitioning
between liquid and solid precipitation events at a site, important changes in
rainfall dynamics are expected, especially in mountainous areas. Thus, an
important issue for modern hydrology is to determine how climate change would
affect the liquid-solid partitioning of precipitation and its statistical
properties. The main aim here is to determine, via statistical analysis and
goodness-of-fit tests, whether the duration of precipitation events under the
different forms (namely solid, liquid and mixed) may be characterized by the
same probability distribution. Similar issue is tested for the volume of
precipitation. For this aim, our study pays attention to hourly data collected
along an altitude gradient identified through six automatic weather stations in
Trentino region, northeast Italy. To distinguish the different types of events
from observed heated pluviometers’ data, a partitioning procedure has been used
and validated, through some disdrometer data. Sample data of duration and
volume, relatively to solid, liquid and mixed events, are extracted, and
univariate and bivariate statistics are calculated. Then, the two-sample
Kolmogorov-Smirnov test is used to test if the data distinguished by different
types of precipitation can be considered extracted from the same distribution.
The results showed that in most cases, durations, as well as volumes of the
different types of events, cannot be considered equally distributed. This
consideration is particularly clear at high elevations.