Abstract

CAMPOS-ARANDA, Daniel Francisco. Detection of homogeneous records in 16 large series of annual precipitation of the Potosino Plateau, Mexico. Tecnol. cienc. agua [online]. 2020, vol.11, n.3, pp.107-157.  Epub June 10, 2024. ISSN 2007-2422.  https://doi.org/10.24850/j-tyca-2020-03-04.

Homogenization methods have been developed to reduce the impact of non-climatic alterations in the records or series of data coming from climatological stations. Such impacts arise from changes either in the location of the station, in the equipment used or in the measurement technique, as well as from alterations suffered by the surroundings. Due to the fact that the historic information on the changes that occurred in the climatological station is generally not available, statistical tests are used to detect breaking points that imply changes in the mean and, therefore, loss of homogeneity of the record. There are two approaches to the application of statistical tests: (1) individually to each record to be tested and (2) using reference series based on neighboring stations. In this study, the first approach was applied to a specific geographical area and using the following four tests: Von Neumann, Pettitt, Buishand and SNHT. Sixteen annual precipitation series of the Potosino Plateau of Mexico were processed in the common period from 1964 to 2016, the amount of values varied from 42 to 53 data. Monthly data were used; hence, the procedures to deduce missing values and to detect and correct extreme maximums are exposed. A description of how the missing annual values were estimated is also included. It was found that 11 records were reliable, because no test detected them as non-homogeneous. Three records were classified as less reliable, because a maximum of two tests found them non-homogeneous and finally, two records are unreliable, since a minimum of three tests find them non-homogeneous. The adopted procedure proposes a practical strategy to detect homogeneous records of annual precipitation; its advantages are exposed at the study.

Keywords : homogenization; maximum outliers values; von Neumann test; Pettitt rank test; Buishand rank test; SNHT test; monthly and annual rainfall.