One-way ANOVA test of five digital filter recursive graphic methods in baseflow separation on Wae Tomu Watershed Ambon City

Recursive digital hydrographic is an automatic and recommended method of separating baseflow to obtain reproducible results. Digital filters help assess the contribution of water to the total flow of a river, while measuring the flow of the riverbed requires sensitive and accurate flow separation techniques. Therefore, this study aims to analyze the analysis of one-way ANOVA digital filter methods for the hydrological characteristics of the small island watershed of Ambon City, including RDF one parameter algorithm, two-parameter algorithm, RDF Lyne & Holick, RDF Chapman, and RDF Eckhard. The one-factor Fingerprint Analysis (ANOVA) results for the five RDF methods of baseflow separation showed that the distribution of baseflow data was generally with a Shapiro-Wilk coefficient greater than α 0.05 and homogeneity with a Levene coefficient of 0.560 greater>α 0.05. The baseflow separation of the five RDF methods has a significant difference with a significance value of 0.001<α 0.05, and the highest average base flow value of the five RDF methods in a row is the Two-parameter algorithm method, followed by the One-parameter algorithm, Eckhardt algorithm, Lyne &Hollick algorithm, and Chapman algorithm. In conjunction with the Turkish test, this calculation showed that the flow separation for the four RDF methods, Lyne &Hollick algorithm, Eckhardt algorithm, One-parameter algorithm, and two-parameter algorithm, have significant similarities and differ significantly from the RDF-Chapman method. These results are strongly influenced by the hydrological characteristics of small island watersheds with limited rain catchment areas and are vulnerable to global environmental changes.