Estimation of daily reference evapotranspiration across different Brazilian climatic conditions, with limitation of meteorological data
Name: IULO PESSOTTI MORO
Publication date: 23/03/2026
Examining board:
Name |
Role |
|---|---|
| ALEXANDRE CANDIDO XAVIER | Examinador Externo |
| ELIAS FERNANDES DE SOUSA | Examinador Externo |
| JOSE EDUARDO MACEDO PEZZOPANE | Examinador Interno |
| MARCEL CARVALHO ABREU | Examinador Externo |
| SIDNEY SARA ZANETTI | Presidente |
Summary: Efficient water resource management is a key challenge for agricultural and forestry sustainability in Brazil, given the country's continental dimensions and climatic diversity. Accurate estimation of reference evapotranspiration (ETo) is essential for irrigation management and hydrological studies, with the Penman-Monteith FAO-56 (PM) method being the internationally recommended standard. However, the application of this method in Brazil is hampered by the scarcity of meteorological data and the low density of stations. This thesis conducts exploratory studies and proposes methodological solutions to better estimate ETo, adapting models to the national climate reality. To this end, a meteorological database from 584 automatic weather stations of the National Institute of Meteorology (INMET), distributed throughout Brazil, was used, including maximum and minimum air temperature, maximum and minimum relative air humidity, global solar radiation, and wind speed. The historical series covers the period from the start of station operations (the oldest dating back to 2000) to 2023, totaling an average of 14.6 years of daily data at each weather station used. In Chapter 1, an initial diagnosis was performed, comparing 12 empirical methods with the FAO-56 PM standard. The results show that, among the temperature-based methods, Hargreaves-Samani performed best. However, the indiscriminate use of simplified (empirical) equations can generate systematic errors that vary according to the climate zone, compromising water resource management, especially when applied without regional calibration. Chapter 2 presents an investigation through a sensitivity analysis of the Penman-Monteith equation. The weight of each meteorological element in the ETo estimate was identified, demonstrating that while solar radiation is more important in humid regions, wind and relative humidity are predominant in semi-arid climates. Chapter 3 proposes an unprecedented correction for situations where relative humidity data is missing. The dew point temperature was adjusted based on the minimum air temperature, also using the daily temperature range (AT). This method allowed for a more efficient estimation of actual vapor pressure, also enabling better daily estimates of ETo using the FAO-56 PM method, even with missing data. Finally, Chapter 4 addresses the most restrictive scenario, where only air temperature is available. The calibration of the Hargreaves-Samani equation based on AT proved to be superior to the PM method based only on temperature, offering a more robust alternative for regions with low data availability. Thus, this thesis seeks to contribute through a set of adapted methodological tools aimed at providing greater security in water resource management, given the limitation of available meteorological data.
Keywords: Penman-Monteith FAO-56, Hargreaves-Samani, temperature range, incomplete weather series, dew point temperature.
