All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 140 10 0
PDF Downloads 3 0 0

On the Extension of Climatic Series from Short Records

View More View Less
  • a Department of Meteorology, Universidad de Buenos Aires, Buenos Aires, Argentina
  • | b Centro Nacional Patagónico, Consejo Nacional de Investigaciones, Científicas y Técnicas, Buenos Aires, Argentina
Restricted access

Abstract

A method is described for finding the minimum span of data required for climatic series extension to a specified and longer period within a RMSE bound, using a principal components regression technique (PCT).

The mean square error (MSE) of a PCT extended series is derived showing that the MSE depends on a term due to the truncation in the eigenvector representation and on another term due to the PCT extension of a short record of data.

The method is applied to four sets of seasonal precipitation series from four different climatic regions of the world. Results show that series extensions can be made within expected RMSE bounds from records which in some cases are shorter than 20% of the total extended period. The PCT series extensions give substantially more information than seasonal averages, and they are also better than series extensions done with simpler techniques.

Abstract

A method is described for finding the minimum span of data required for climatic series extension to a specified and longer period within a RMSE bound, using a principal components regression technique (PCT).

The mean square error (MSE) of a PCT extended series is derived showing that the MSE depends on a term due to the truncation in the eigenvector representation and on another term due to the PCT extension of a short record of data.

The method is applied to four sets of seasonal precipitation series from four different climatic regions of the world. Results show that series extensions can be made within expected RMSE bounds from records which in some cases are shorter than 20% of the total extended period. The PCT series extensions give substantially more information than seasonal averages, and they are also better than series extensions done with simpler techniques.

Save