The efficiency of average linkage hierarchical clustering algorithm associated multi-scale bootstrap resampling in identifying homogeneous precipitation catchments

Identifying the Ideal Number Components of the Bayesian Principal Component Analysis Model for Missing Daily Precipitation Data Treatment

“…In contrary, the BPCA5-VB algorithm is more superior compared to the single imputation data algorithms proposed in literature, which ranked in the best three irrespective of missing rates. Due to the spatial and temporal characteristics between the daily precipitation time series data acquired from the monitoring stations located in inland and coastal region are varied [6], therefore a vary imputation algorithm is indeed much needed in treating the missing daily precipitation time series. Mean 6 7 7 7 7 7 Median 3 2 6 6 6 6 BPCA2-VB 7 1 3 5 4 1 BPCA3-VB 4 5 5 4 1 5 BPCA4-VB 5 6 4 1 2 4 BPCA5-VB 2 4 1 2 3 3 BPCA6-VB 1 3 2 3 5 2 3732020 Mean 7 7 7 7 7 7 Median 1 6 6 6 6 6 BPCA2-VB 2 4 5 1 4 4 BPCA3-VB 6 5 4 4 5 2 BPCA4-VB 4 3 3 3 1 3 BPCA5-VB 3 2 1 5 2 1 BPCA6-VB 5 1 2 2 3 5…”

“…This district is frequently exposed the risk of occurrence for extreme precipitation tragedies when the monsoon season is prevailing. In this study, daily precipitation time series of monitoring stations located in the inland and coastal regions in the Kuantan district [6] as depicted in Fig. 1 are selected to evaluate the effectiveness of the proposed multiple imputation algorithm, which this daily precipitation time series data are acquired from the Department of Irrigation and Drainage (DID) Malaysia.…”

A review on homogeneity across hydrological regions

“…Latt et al [3] presents a great overview from a statistical viewpoint. Chuan et al [29] define eight popular algorithms in order to make suitable hierarchical clustering of rivers in regards to their flow regimes:…”

“…Hierarchical and flat clustering techniques may apply together to develop hydrologic categorisation (called hybrid clustering). For instance, Chuan et al [29] employed a partitioned clustering progression in order to distinguish groups of alike catchments through improving the clusters obtained from "agglomerative hierarchical clustering algorithms" applying the k-means algorithm. Also, Belletti et al [30] examined outputs of a hierarchical, average-connection algorithm to assist distinguish an ideal amount of clusters for succeeding flat classification applying k-means.…”

“…Chuan et al [29] have provided clear description of hard clustering algorithms in regionalization. The most basins can be similar to various basins and thus, assigning a basin to one area (cluster) or another possibly cannot be justified.…”