Incorporação do impacto da rede de reservatórios superficiais artificiais de caráter intranual na modelagem hidrológica chuva-vazão

Abstract

A hydrologic model’s ability to describe the process- es encompassing the transformation of precipitation into flow rates depends on the assumptions, structure and for- mulations it uses. Models of the spatially distributed type are very attractive. However, the mathematical simplifica- tions used may generate errors when the represented basin's features are very heterogeneous due to natural characteris- tics or to anthropic activity. The river regime in the semi- arid Northeast of Brazil has undergone changes due to the construction of superficial artificial reservoirs, most of which are small reservoirs of an intra-annual nature. Those small reservoirs make modeling difficult. In this paper we propose a change to the structure of the SMAP model aiming to explicitly incorporate the represen- tation of small damming. This representation is done by inserting a reservoir into the mathematical model with the operating characteristics which are representative of small damming behavior. The dimension of that reservoir is a gradable parameter (h RPA) aiming to identity the effect of this intra-annual regularization. The proposed model, SMAP-RPA, consists of expanding the SMAP model in its monthly version. Therefore, determin- ing the results from the RPA component improves the mod- el’s performance and makes it more realistic when compared to SMAPm's performance. As a result, in cases where the dams in the hydrographic basin do not produce significant variation in the hydrograph, the “hRPA” parameter is null and the SMAP-RPA operation is the same as in SMAPm. Aiming to ascertain the proposed model's validity, we de- veloped a case study by formulating two scenarios accord- ing to the flow rate series and the characteristics of 18 hy- drographic basins located in the state of Ceará. The results support the efficiency of small damming representation and the ensuing efficiency improvement in the adjustments to observed and calculated series. In this paper we have also noticed an important scale effect. The basins which were affected most by small damming were those with a drainage area of less than 5,000 km2. The impact of this small damming on larger scale basins is not detected by the mod- el, which makes the “h RPA” parameter null