Forecasting glacier snout fluctuations in the western italian Alps.
Abstract
Glaciers in the western italian Alps show a long-term (centennial) widespread retreat, modulated by fluctuations on decadal time scales, lagging the variability of the regional temperature-precipitaion regime. The lag of maximum correlation is 8 years for temperature and 10 years for precipitation, which is considerably shorter than the response-time estimates based on linearized dynamics. We propose an empirical model aimed at establishing a quantitative relationship between the variability of the terminus positions of a set of selected glaciers in North-Western Italy and the corresponding climatic fluctuations, namely of precipitation rates and mean temperatures. The link is sought in the form of a stochastic lagged linear model. Linearity in glacier fluctuation response is assumed as the simplest working hypothesis. The stochastic component of the model aims at accounting for the unexplained dynamics that is always present in natural systems with a large number of degrees of freedom. As usual, we require the stochastic term to be gaussian and uncorrelated. Surrogate data techniques based on the generation of synthetic datasets are used for robust estimation of the model parameters and hyptesis testing. The linear stochastic model describes about 66% of the variance of the average snout fluctuation and precipitation appears to give the major contribution to the explained variance. Using temperature and precipitations data from a large set of local wheather stations, the average trend for the snout of the glaciers is forecasted for the next.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2003
- Bibcode:
- 2003AGUFM.C22B..03C
- Keywords:
-
- 1827 Glaciology (1863);
- 3354 Precipitation (1854)