Banca de DEFESA: GEIZA THAMIRYS CORREIA GOMES

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : GEIZA THAMIRYS CORREIA GOMES
DATE: 24/02/2023
TIME: 14:00
LOCAL: videoconferência
TITLE:

EVALUATION OF MONTHLY CLIMATE PRECIPITATION FORECASTS IN THE MUNDAÚ RIVER RIVER BASIN VIA STATISTICAL DOWNSCALING OF THE GLOBAL CLIMATE MODEL METEO-FRANCE -SYSTEM 7

KEY WORDS:

Weather forecasts, monthly cumulative precipitation, CHIRPS, BHM, MFS7.

PAGES: 80
BIG AREA: Ciências Exatas e da Terra
AREA: Geociências
SUBÁREA: Meteorologia
SUMMARY:

Producing sub-seasonal climate forecasts is crucial for many economic sectors and has great relevance for society as a whole. Different climatic events act in the Northeast region of Brazil, conditioning success or failure in agricultural activities, recharge or shortage of water resources, periods of dry or humid extremes. The Mundaú River Basin (BHM) is one of the most important for the states of Alagoas and Pernambuco, with a tropical/semi-arid climate where the watercourse and territorial extension of the basin crosses and divides the two states. In this basin there are cyclic occurrences of long periods of drought and severe floods. Faced with this problem, the objective of the present study was to regionalize the climate forecasts of the French model Méteo-France System 7 (MFS7) for the BHM using the Canonical Correlation Analysis (ACC) technique, which allows to recalibrate the climate forecasts retrograde measurements of a model by confronting them with observations in an area, and evaluating its dexterity afterwards. The forecast for each month was obtained up to five months in advance in the period 1993-2016 and analyzed deterministically through the correlation between simulated and observed values, showing as the main result that the forecast for a given month was carried out exactly one month in advance. , it generally provides the most accurate forecast, although this is not a rule for every month of the year. However, it was noticed that when applying an average forecast per sets, from the average of all five previous predictions, the correlations are significantly increased between this average prediction and the observations, in addition to the decrease in errors related to the prediction. The network of observations in the BHM is not extensive and the source of the observations used was from the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS), which proved to be efficient for estimating the monthly accumulated rainfall in the BHM when compared to the few observed series. After evaluating the recalibrated forecasts with ACC, a case study was carried out applying the set forecast for all months of the year 2020, with results animators who indicated climate predictions consistent with the observations of this same year, demonstrating the operational potential of using the MFS7 climate data to generate reliable climate forecasts for the BHM.

BANKING MEMBERS:
Presidente - 1537309 - FABRICIO DANIEL DOS SANTOS SILVA
Interno(a) - 1846078 - GLAUBER LOPES MARIANO
Externo(a) à Instituição - JONATHAN MOTA DA SILVA - UFRN
Interno(a) - 1653612 - MARIA LUCIENE DIAS DE MELO