Department of Civil and Environmental Engineering, Federal University of Paraíba, João Pessoa, PB, 58051-900, Brazil.
Department of Geosciences, Federal University of Paraíba, João Pessoa, PB, 58051-900, Brazil.
Environ Monit Assess. 2018 Feb 26;190(3):175. doi: 10.1007/s10661-018-6536-3.
Trend analysis is an important issue for the decision-making processes. Thus, trends of rainfall, consecutive dry days (CDD), and consecutive wet days (CWD) in the Upper São Francisco River basin, Brazil, using daily rainfall data from the Tropical Rainfall Measuring Mission (TRMM) for recent 18 years, were analyzed. Instead of analyzing the trend of one average time series for one specific confidence level, a spatiotemporal analysis over the entire area with 169 continuous time series is done by applying the nonparametric Mann-Kendall and Sen tests for simultaneously 13 confidence levels and a new integrated confidence classification is proposed. The results show that the rainfall has increased during the less rainy periods (from June to October) and has decreased in the rainy periods (from November to May), with the highest and lowest confidence levels, respectively. An analysis of CDD and CWD shows that the number of CDD has decreased, while the number of CWD has increased, which revealed that the dry periods are more frequently interrupted for the period studied.
趋势分析是决策过程中的一个重要问题。因此，使用来自热带降雨测量任务（TRMM）的最近 18 年的每日降雨数据，对巴西上圣弗朗西斯科河流域的降雨、连续干燥天数（CDD）和连续湿润天数（CWD）的趋势进行了分析。本研究没有分析一个特定置信水平的一个平均时间序列的趋势，而是通过应用非参数 Mann-Kendall 和 Sen 检验对整个区域的 169 个连续时间序列进行时空分析，同时提出了一种新的综合置信度分类方法。结果表明，在较干燥的时期（6 月至 10 月），降雨量增加，而在雨季（11 月至 5 月），降雨量减少，分别具有最高和最低的置信度水平。对 CDD 和 CWD 的分析表明，CDD 的数量减少，而 CWD 的数量增加，这表明在研究期间干燥期更频繁地被打断。
