Defense
Student: Eliana Nascimento da Costa
Program: Meteorology
Title: "Distribution of raindrop size at the ATTO-Campina experimental site"
Advisor: Profa. Dra. Rachel Ifanger Albrecht
Judging Comitee:
- Profa. Dra. Rachel Ifanger Albrecht - Chair and Advisor - IAG/USP
- Dra. Izabelly Carvalho da Costa - CPTEC/INPE (by videoconference)
- Dr. André Cezar Pugliesi da Silva - Post-doc - IAG/USP
Other Members:
- Prof. Dr. Micael Amore Cecchini - IAG/USP
- Dr. Thiago Souza Biscaro - CPTEC/INPE
- Prof. Dr. Rafael Castelo Guedes Martins - UFERSA
Abstract:
This study presents a characterization of the raindrop size distribution (DSD) at the Amazon Tall Tower Observatory site (ATTO-Campina) in central Amazonia, based on Joss- Waldvogel impact disdrometer (RD-80) measurements from January 2020 to August 2025. A total of 122,167 one-minute samples with R ≥ 0.1 mm h−1 were analyzed across the wet season (79% of samples), wet-to-dry transition (6%), and dry season (15%). DSDs were fitted with normalized gamma distributions, and convective/stratiform (C/S) classification followed a log10(Nw) = 3.85 threshold proposed in the literature for tropical regions. The log10(Nw) distribution is bimodal in all seasons, with modes near 3.4 and 4.0 during the wet season and near 1.95 and 3.8 during the dry season, indicating the coexistence of convective (warm rain) and stratiform precipitation. Overall, 34.4% of the samples are classified as convective and account for 55.2% of accumulated rainfall; this fraction rises to 61% in the wet season and drops to 33% in the dry season. The result that best characterizes ATTO-Campina microphysics in the tropical context is the convective regime signature: convective log10(Nw) (4.17) matches equatorial Indian Ocean reference values, yet convective Dm is slightly smaller (1.04 mm versus 1.14 mm) and mean convective R is considerably higher (7.37 versus 5.34 mm h−1). This combination (similar Nw, smaller drops, more intense rainfall) is consistent with the hypothesis that continental Amazonian warm rain may be more efficient at converting cloud water into precipitation than its oceanic warm-pool counterpart. In the Z = A RB relations, the convective prefactor (A ≈ 110 ) is about three times smaller than the stratiform one (A ≈ 330 ), indicating that a single Z–R relation tends to underestimate convective rainfall and overestimate stratiform rainfall. Two empirical orthogonal function (EOF) modes explain over 94% of the total variance, indicating that the DSD can be described with few parameters. The five-year ATTO-Campina record represents one of the longest disdrometer-based DSD datasets in central Amazonia and provides regional parameters that may contribute to radar-based precipitation estimation and to microphysical parameterizations in atmospheric models.
Keywords: drop size distribution, disdrometer, central Amazon, ATTO, Z-R relation, precipita- tion microphysics.