Banca de DEFESA: AMANDA RODRIGUES DE SOUZA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : AMANDA RODRIGUES DE SOUZA
DATE: 07/12/2022
TIME: 14:00
LOCAL: Remota
TITLE:

PRECIPITATION INDICES FOR STREAMFLOW PREDICTABILITY IN BRAZILIAN HYDROGRAPHIC BASINS

PAGES: 155
BIG AREA: Ciências Exatas e da Terra
AREA: Geociências
SUBÁREA: Meteorologia
SPECIALTY: Climatologia
SUMMARY:

Variability in precipitation can cause streamflow anomalies, and depending on the intensity and persistence, lead to the occurrence of extreme cases, which can affect the availability of water resources and socioeconomic activities. The objective of the study is to investigate the influence of precipitation on the streamflow, at points of hydroelectric use, seeking to identify precipitation indices with a potential predictor of streamflow. Rainfall data from ANA and streamflow data from ONS and ANA were used for the period from 1980 to 2020. However, four streamflow points were focused: Jurumirim, Porto Primavera, Sobradinho and Tucuruí. The precipitation stations upstream of the streamflow points were grouped by cluster analysis into homogeneous precipitation regions. The Jurumirim flow point presented a region with 17 stations. Porto Primavera had eight seasons, where the regions had 68, 3, 7, 141, 1, 37 and 240 seasons. Sobradinho presented five regions, where the regions presented 4, 2, 10, 65, 20 stations. Tucuruí presented six regions, where the regions presented 17, 8, 66, 5, 40 and 3 stations. Precipitation indices were calculated, namely: number of days with precipitation above the quantile thresholds of 85% and 65% (ND85q+ and ND65q+), below the quantile thresholds of 35% and 15% (ND35q- and ND15q-), accumulated precipitation on wet days (PRCTOT), simple daily intensity indices (SDII), average and maximum sequences of days with and without precipitation (CWD – MED, CWD – MAX, CDD – MED and CDD – MAX), count of days without (ND1mm -) and with precipitation above 1 mm, 10 mm, 20 mm and 30 mm (R1mm, R10mm, R20mm and R30mm) and the accumulated value of precipitation in one, three, five, seven and ten consecutive days ( RX1DAY, RX3DAY, RX5DAY, RX7DAY and RX10DAY). A simple correlation analysis was performed between the streamflow series and the precipitation indices with annual lags of up to two years and monthly lags of up to twelve months. Different sets of indices were identified for each of the precipitation regions, highlighting the differences in the behavior of precipitation influence. The results suggested the use of movable annual indices for streamflow prediction, which were applied to multiple linear regression models. The best models were obtained for a lag of one month, especially Sobradinho, for presenting the highest value of the coefficient of determination. For the prediction in Jurumirim, only the one-month lag model presented a precipitation index as a predictor, being the PRCTOT index. For Porto Primavera, precipitation indices were selected from the 1PP regions (PRCTOT), the 3PP region (CWD – MED) and the 7PP region (ND85q+ and ND35q+). For Sobradinho, precipitation indices were selected from the 2SF regions (PRCTOT, CWD – MED and ND35q-), the 3SF region (ND35q- and ND85q+) and the 4SF region (ND65q). For Tucuruí, precipitation indices were selected from the 2T regions (PRCTOT, CDD – MED and ND65q+) and the 6T region (ND85q+ and ND35q-). The results indicate that the indices are promising in the predictability of the streamflow and can be used in a complementary way to climatic indices.

BANKING MEMBERS:
Presidente - Interno ao Programa - 1809834 - ANDREA DE OLIVEIRA CARDOSO
Membro Titular - Examinador(a) Externo ao Programa - 1544340 - PATRICIA TEIXEIRA LEITE ASANO
Membro Titular - Examinador(a) Externo à Instituição - SIMONE EROTILDES TELEGINSKI FERRAZ - UFSM
Membro Suplente - Examinador(a) Interno ao Programa - 1876379 - MARIA CLEOFE VALVERDE BRAMBILA
Membro Suplente - Examinador(a) Externo à Instituição - LUANA ALBERTANI PAMPUCH - UNESP