Forecasting local daily precipitation patterns in a climate change scenario

The present study introduces a statistical procedure for obtaining long-term local daily precipitation forecasts in a climate change scenario. It is based on a regression model that uses climate variables properly reproduced by a General Circulation Model (GCM) as predictors. The daily rainfall mode...

Full description

Saved in:
Bibliographic Details
Published in:Climate research 2005-05, Vol.28 (3), p.183-197
Main Authors: Abaurrea, Jesús, Asín, Jesús
Format: Article
Language:English
Subjects:
Atmospheric models
Climate change
Climate models
Forecasting models
Humidity
Modeling
Precipitation
Rain
Seasons
Weather forecasting
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The present study introduces a statistical procedure for obtaining long-term local daily precipitation forecasts in a climate change scenario. It is based on a regression model that uses climate variables properly reproduced by a General Circulation Model (GCM) as predictors. The daily rainfall model used consists of a logistic regression as the occurrence model and a generalized linear model (GLM) with Gamma error distribution as the quantity model. The ability of the model to generate plausible long-term projections is analysed by studying and comparing its behaviour using observed and GCM simulated data as input. The method is applied to forecast the rainfall pattern in the area of Zaragoza (Spain) for the period 2090–2100, in an IS92a scenario. We use the data corresponding to an experiment with the CGCM1 model, the first version of the coupled GCM of the Canadian Centre for Climate Modelling and Analysis (CCCma). The results obtained show that no significant change in global rainfall frequency or in the annual accumulated amount are to be expected; however, an important modification of the seasonal cycle, with a high decrease in rainfall frequency and in the amount collected in spring, is forecasted.
ISSN:0936-577X
1616-1572
DOI:10.3354/cr028183