This paper presents the second part of the methodology developed to project hydrometric networks of quantitative monitoring of watersheds, emphasizing exclusion, inclusion and rearrangement of monitoring stations. The presented methodology has as study area the Doce River Basin (DRB) and it is based on spatial analysis of maps generated by kriging for regionalized variables: minimum specific flow of 7 days duration and return period of 10 years - q7,10 and precipitation of dry season - Pss. Applying the methodology developed for the DRB, it was concluded that there were significant improvements in the standard deviations of estimates of flow for 6 of the 9 ottobacias, culminating with the rearrangement of 18 stations and excluding 2 stations in the gauging monitoring network. It was also observed that there were improvements too in the estimates of spatialized precipitation enabling the spatial rearrangement of 12 stations. It is noteworthy, however, the need for inclusion of 67 stations in the network monitoring rainfall. The methodology was adequate for the proposed objective and can be applied to any watershed.
geostatistical analysis; fluviometric monitoring network; rainfall monitoring network
