Water scarcity in Brazil: part 1—regionalization of the AWARE model characterization factors

Purpose This paper presents the regionalized water scarcity characterization factors (CFs) of the available water remaining (AWARE) model, which was found by a previous study, on the water scarcity in Brazil, to be the most indicative characterization model for the water-scarce regions in Brazil. We used the national database and hydrographic delimitations defined by the National Water Agency (Agência Nacional de Águas — ANA) to generate the regionalized AWARE BR CFs. Methods The CFs were regionalized by hydrographic delimitations used by ANA: (i) State Hydrographic Units (SHU) and (ii) Hydrographic Regions (HR). These AWARE BR CFs were compared with the factors originally proposed by WULCA ( 2018 ) and with the Scarcity Index used by ANA to identify the scarcest regions in the country. Finally, the AWARE and AWARE BR factors were applied to a case study of Brazilian melons, evaluating the regionalization effects on the results of water scarcity analysis. Results and discussion The AWARE BR CFs demonstrate most consistency with the regions recognized by ANA to have water scarcity problems, such as the semiarid region. Approximately 12% of the SHUs exhibited maximum water scarcity (CF = 100) during the entire year, while 11% presented minimum scarcity factors (CF = 0.1). The comparison of hydrologic data from ANA with those from WaterGAP indicated that water availability was overestimated in WaterGAP, while demand was underestimated in different basins. The comparison of AWARE BR CFs with ANA Scarcity Index values indicated more similarity (smaller residual error) than the comparison of AWARE BR CFs with AWARE. The case study regarding the impact of water scarcity on melons showed a significant difference between characterization factors and, consequently, in the values of impact. Conclusions AWARE BR factors generated with national characterization data are adapted to the different regions of Brazil, exhibiting higher sensitivity to the semiarid region. This regionalization provided a more accurate representation of the scarcity in smaller basins located in larger basins, characterized by large climate variation.