SATELLITE RAINFALL RETRIEVAL s VALIDATION: Spatial distribution OF statistical
parameters during winter and summer season
Wagner Flauber A. Lima1 ; Éder Paulo Vendrasco1 ; Daniel Alejandro Vila1,2
1Centro de Previsão de Tempo e Estudos Climáticos – CPTEC/INPE
2CICS/ESSIC-NOAA, University of Maryland College Park
[email protected]
ABSTRACT
This work examines different methods for estimating rainfall over South America during winter and summer season for the period 2009-2011. The algorithms
analyzed in this study are: Hidroestimator, 3B42RT, CMORPH, and GSMAP. The evaluation of the algorithms was performed by comparing them with rain
gauges data. The analysis showed that 3B42RT and CMORPH methodologies got better performance than others for the studied period. Another important
observation was the poor performance in estimating precipitation in northeastern Brazil.
RESULTS
INTRODUCTION
Modelos
Correlação - Verão
Precipitation is one of the most important atmospheric components of the
earth-atmosphere system as it affects the environment in various ways and
thus society. It is among the most difficult parameters to measure, due to its
high variability in time and space. One of the main ways to get information
about the precipitation on a global scale is using satellite-based rainfall
retrieval techniques.
Correlação - Inverno
The aim of this study is to evaluate different methods for estimating daily
rainfall obtained from different sources: Hidroestimador (Vicente, 1998),
3B42RT (Huffman et al., 2003), CMORPH (Joyce et al., 2004) and GSMAP
(Aonashi et al ., 2009) over South America for the summer and winter
seasons of the period 2009-2011.
Figura 1 - Distribuição espacial da correlação entre a precipitação observada e a
estimada para o inverno (base) e verão (topo) dos anos de 2009, 2010 e 2011.
Modelos
POD - Verão
DATA AND METHODOLOGY
For the development of this work, we use daily totals from different
regional networks and synoptic stations reported regularly by the "Global
Telecommunication System" (GTS), interpolated on a grid of 0.25 ° latitude
and longitude over South America and daily accumulated precipitation from
four rainfall estimation models:
POD - Inverno
 3B42RT, CMORPH, GSMAPT e HIDROESTIMADOR;
 Winter and summer season for the period 2009-2011.
Typical parameters for assessing rainfall retrievals are: Probability of
Figura 2 - Distribuição espacial da Probabilidade de detecção (POD) dos modelos de
estimativa de precipitação para o inverno e verão dos anos de 2009, 2010 e 2011.
Detection (POD), False Alarm Ration(FAR) and correlation (COR). Assuming
Modelos
FAR - Verão
N the total number of cases and a, b,​ c and d set for two data sets A (model)
and B (gauge) as follows:
a
a
POD =
a+c
;
Positive hit
Precipitation in A & B
b
False Alarme
Precipitation in A only
c
Surprise
Precipitation in B only
d
Positive
Negative
No precipitation
b
FAR =
a+c
;
COR =
FAR - Inverno
∑( A − A ).( B − B )
∑( A − A ) .∑( B − B )
i
i
2
i
Figura 3 - Distribuição espacial do Falso Alarme (FAR) dos modelos de estimativa de
precipitação para o inverno e verão dos anos de 2009, 2010 e 2011.
i
RESULTS
2
The models showed that the best performances were for 3B42RT and
CMORPH, mainly to the Southern and Southeastern Brazil. Another
important fact was the poor performance of the modelsin the northeastern
region, in both seasons.
 Analyzing the fields of correlation (Fig. 1) for the four models, the best
performances were obtained for the winter season, especially in the South
and Southeast of Brazil.
References
 It is observed that the best values of
​ POD (Fig. 2) are obtained for the
summer season mainly in northern Brazil for all models. However, the best
results were obtained for the model 3B42RT.
Joyce J.,J. Janoviak,P. Arkin,P. Xie: CMORPH:A Method that Produces Global Precipitation Estimates from
Passive Microwave and Infrared Data at High Spatial and Temporal Resolution. J.Hydrom.,5,487–503,2004.
 The models showed low values for FAR (fig. 3) during the summer
season, especially in Southern and Southeastern Brazil. In general, rainfall
retrieval models show similar results in both seasons for FAR index.
Vicente, G. A.: The operational GOES infrared rainfall estimation technique. Bulletin of the American
Meteorological Society, 79(9), 1883-1898, 1998.
Huffman, G.J., R.F. Adler, E.F. Stocker, D.T. Bolvin, and E.J. Nelkin, 2003: Analysis of TRMM 3-Hourly MultiSatellite Precipitation Estimates Computed in Both Real and Post-Real Time. Combined Preprints CD-ROM,
83rd AMS Annual Meeting, Poster P4.11 in: 12th Conf. on Sat. Meteor. and Oceanog., 9-13 February, Long
Beach, CA, 6pp, 2003.
K. Aonashi, J. Awaka, M. Hirose, T. Kozu, T. Kubota, G. Liu, S. Shige, S. Kida, S. Seto, N. Takahashi, and Y. N.
Takayabu, 2009: GSMaP passive, microwave precipitation retrieval algorithm: Algorithm description and
validation. J. Meteor. Soc. Japan, 87A, 119-136.
The authors acknowledge Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ) for
financial assistance (Process nº 476599/2010-5).