Frontiers in Digital Humanities | |
Variability, Statistical Distributions and Trends of Precipitation Extremes on the Island of Sardinia, 1951â2000: Result Dependence on Network Density | |
Boi, P.1  | |
[1] Dipartimento Meteoclimatico, ARPAS Sardegna, Italy | |
关键词: Rainfall extremes; Sardinia; Trends; Network density; Gumbel distribution; Mann-Kendall test; variability; | |
DOI : 10.3389/feart.2018.00188 | |
学科分类:社会科学、人文和艺术(综合) | |
来源: Frontiers | |
【 摘 要 】
Abstract. An analysis of high resolution network data of daily precipitation on Sardinia island from 1951 to 2000 is presented. The mean distance between the stations is about 10 km. Comparison between monthly series shows significantly higher extreme values in October and November. Decadal mobile window Wilkoxon-Mann-Whitney test indicates that the decade 1965-75 have significantly higher values than the rest of yearly series, while the decade 1976-85 has lower values. Trends by Mann-Kendall test present not significant results except March (decreasing) and July (increasing). The monthly and yearly series are fitted by the Gumbel distribution. The dependence of the previous results on the network density has been investigated. A second set of data has been selected by eliminating about 20% of stations in a uniform way, so the station mean distance raises to about 12 km; then a third set of data has been obtained by eliminating about 80% of stations, so the station mean distance is 20km. The previous analysis have been repeated with the two other sets. Comparison between the original data set and the second one by Wilkoxon-Mann-Whitney test shows that they are very similar, on the contrary the third one is very different. The fitting distributions of the original data set and the second one are almost identical, while the fitting distributions of the third data set are very different. Also different results are obtained for the third data set in the trend test.
【 授权许可】
CC BY
【 预 览 】
Files | Size | Format | View |
---|---|---|---|
RO201904025048592ZK.pdf | 6909KB | download |