Banca de DEFESA: MARCELLA VASCONCELOS QUINTELLA JUCÁ

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : MARCELLA VASCONCELOS QUINTELLA JUCÁ
DATE: 17/08/2021
TIME: 09:00
LOCAL: Auditório Pós-Graduação Engenharia Civil - CTG - UFPE
TITLE:

SOIL MOISTURE ESTIMATION AT DIFFERENT DEPTHS FROM THE SOIL MOISTURE OCEAN SALINITY (SMOS) SATELLITE DATA IN PERNAMBUCO USING AN EXPONENTIAL FILTER

KEY WORDS:

Soil Moisture. Exponential Filter. SMOS.

PAGES: 75
BIG AREA: Engenharias
AREA: Engenharia Civil
SUMMARY:

Droughts are natural disasters whose severity and duration are often hard to predict. This extreme event can be identified, for example, when soil moisture contents lower than average are detected. Soil moisture at subsurface depths plays an important role in ecological and socioeconomic aspects, as agriculture is a business field heavily affected by droughts, given the crops dependence on the availability of water on the root zone, and therefore, of the soil moisture. In this sense, this parameter is used in hydrological and climate models as well as in agricultural monitoring, as it is an indicator for the availability of water for transpiration and provides predictability regarding the occurrence of agricultural droughts. Among the measurement methodologies for this parameter, remote sensing is a viable option to obtain soil moisture information on a large scale, since the data generated by satellite has a large spatial coverage and has a satisfactory temporal frequency. The SMOS mission measures soil moisture information through a radiometer that operates in the L-Band. Therefore, the data obtained refers to the moisture in the first centimeters of the soil. In this study, an exponential filter to estimate the root zone water content from time series of surface moisture obtained via remote sensing was evaluated for in situ stations located in Pernambuco, Brazil based on SMOS data and using in situ stations at 20 cm and 40 cm depths for calibration and parameters setting. The results obtained for the root zone were also compared to the soil moisture products of the GLDAS-Noah and GLDAS-CLSM models. The soil moisture content estimated by the exponential filter resulted in mean Pearson correlation increments and RMSE reductions in the order of 34% and 25%, respectively, when compared to SMOS data without its application, reaching correlations more than 100% higher and RMSE 70% lower and showing the benefits of its application. As for the data resulting from the filter with the GLDAS, a spatial pattern was observed in the comparison between these and the data from the filter.

BANKING MEMBERS:
Externo à Instituição - DANIEL ANDRÉS RODRIGUEZ - UFRJ
Presidente - 1688881 - ALFREDO RIBEIRO NETO
Externa à Instituição - ALZIRA GABRIELLE SOARES SARAIVA SOUZA