Downscaling GCM information to regional scales: a non-parametric multivariate regression approach

Computer Science

Scientific paper

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

22

Scientific paper

The multivariate adaptive regression splines (MARS) model was used to capture the observed relationships between sea level pressure (SLP) anomalies over the Euro-Atlantic sector and the winter time (December February) monthly rainfall at eight sites in Portugal; possible anthropogenic changes of the rainfall in a perturbed future climate were then estimated by using both the observed SLP-rainfall relationships, described by MARS models, and the GCM simulated SLP, taken from the output of the Hadley Centre Transient Climate Change Experiment (UKTR). Also, principal component analysis was carried out to reduce the dimensionality of the SLP data, and to assess the ability of the GCM in simulating the large-scale circulation; the first six principal components were retained as predictors in the MARS model. The MARS model were built up by using the data for 1946 1991 as the training set and that for 1901 1943 as the testing set, showing satisfactory prediction skills. It is concluded that the UKTR control simulation successfully reproduced main features of the large-scale circulation, but the observed relationship between SLP and the regional rainfall was not well preserved. With respect to the 54-year experiment of perturbed run, the MARS estimation of rainfall and the relevant direct GCM rainfall output possess similar multi-annual variations; however, there are substantial differences regarding details; the change in the area mean of winter time mean monthly rainfall in Portugal estimated by MARS (indirect GCM output) is about -12.7 mm per 54-year, and the relevant direct GCM output is -16.9 mm/54-year. This reduction tendency is consistent with previously reported findings respecting rainfall in the Iberian Peninsula, which were based on the MPI (Max-Planck Institute for Meteorology) transient simulations.

No associations

LandOfFree

Say what you really think

Search LandOfFree.com for scientists and scientific papers. Rate them and share your experience with other people.

Rating

Downscaling GCM information to regional scales: a non-parametric multivariate regression approach does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Downscaling GCM information to regional scales: a non-parametric multivariate regression approach, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Downscaling GCM information to regional scales: a non-parametric multivariate regression approach will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFWR-SCP-O-825425

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.