Historical Database for Gridded Daily Precipitation Dataset over South America

Partner Institutions:


The Combined Scheme approach (hereafter, CoSch, Vila et al, 2009) provides a new high-resolution, gauge-satellite-based analysis of daily precipitation over continental South America. This methodology is based on a combination of additive and multiplicative bias correction schemes to get the lowest bias when compared with the observed values (rain gauges). In this case, realtime TRMM Multisatellite Precipitation Analysis (TMPA, Huffman, 2007), (in which no rain gauges are incorporated) is used as a high-quality rainfall algorithm, whereas a daily rain gauge database is used to correct the bias on a daily basis over South America. In a second step, and considering that TRMM mission ended in 2015 and Global Precipitation Measurement (GPM) started in the same year, the replacement of the TMPA algorithm for Integrated Multi-satellitE Retrievals will be considered for GPM (IMERG) once that algorithm can be reprocessed for a suitable number of years to perform a reliable validation.

This project is financed by the InsuResilience Investment Fund (IIF) through the Centro de Gestão de Pesquisa, Desenvolvimento e Inovação (CGPDI), funded by the German Federal Ministry for Economic Cooperation and Development (BMZ) through KfW.

The overall objective of the IIF is to contribute to the adaptation to climate change by improving access to and the use of insurance in developing countries. The specific objective of the Fund is to reduce the vulnerability of poor households as well as micro, small and medium enterprises (MSME) to extreme weather events.


The initial reprocess the entire database from 2000 involve two different tasks:

  • a. Collect and run quality control of global and regional rain gauge data: as described in the Interim Report for the Climate Insurance Fund (CIF), the upgrade of the historical dataset involves two different actions: The first one is to reprocess the current dataset available at CPTEC/INPE from GTS and Brazilian regional networks to include all Latin American countries for the following period: March 2000 – present. A basic QC process was performed.

In order to improve the amount of gauge data, a second action was performed: The Global Historical Climatology Network - Daily (GHCN-Daily) dataset was downloaded and processed to extract daily precipitation data to increase the number of records in the current database and extend the geographical coverage from southern Argentina and Chile to Mexico . GHCN-Daily integrates daily climate observations from approximately 30 different data sources and quality assurance checks are applied to the full dataset.

Figure 1 shows the rainfall field obtained with CoSch for March 3, 2000. All black dots represent the position of raingauges.

  • b. Reprocess daily accumulated satellite-based rainfall fields from 2000 to present: In Mexico, Central and South America, the daily gauge data (24 hours accumulation) is calculated from 12:00 UTC of previous day to 12:00 of the current day. The TMPA algorithm is performed over 3-hour periods for 0.25 degrees spatial resolution. In order to merge both dataset in CoSch, daily accumulations for the same period was done for TMPA-RT. This algorithm was considered because no rain gauges are incorporated. Table 1 summarizes the characteristics of the dataset:
Combined Scheme file characteristics Value
Spatial resolution 0,25°
Spatial coverage Southern Argentina and Chile to Mexico
Temporal resolution Daily
Temporal coverage March 2000 - Present
Geographical Coverage 33 N - 60 S, 120 W - 30 W
Number of lines 372
Number of columns 360
Numeric Format - Generic binary, real*4 , little-endian byte order
- netCDF
Name format CoSch_yyyymmdd1200.bin
File size 535,680 bytes (binary, uncompressed)
To see more information about this project, access this document.


During this project several validations strategies were implemented in order to validate the performance of the LatAm dataset. The first one is a classical statistical approach using a cross correlation process with daily rain gauges as the ground truth, while the second one is based on insurance-relevant losses on the ground across Central and South America, based on tools and information that IRI has developed and gathered through IIF related and other efforts. Using a those tools, it is shown that LatAm dataset is a comparable source to other datasets currently being used for operational index insurance products. More information in ftp://ftp.cptec.inpe.br/rainfall/cif.latam/output_americas/documents/20180501 LatAmDatasetValidation.docx

Data Acess

Next, one can observe the data structure created for the project:

To access project data via ftp, go to the link: ftp://ftp.cptec.inpe.br/rainfall/cif.latam or this link: http://iridl.ldeo.columbia.edu/SOURCES/.INPE/.CPTEC/.latam/.CoSch/