You can download the monthly mean netcdf files from this link:

http://mirador.gsfc.nasa.gov/cgi-bin/mirador/presentNavigation.pl?tree=project&&dataGroup=Gridded&project=TRMM&dataset=3B43:%20Monthly%200.25%20x%200.25%20degree%20merged%20TRMM%20and%20other%20sources%20estimates&version=007&CGISESSID=851b1fb9546fc5ea49699ee2c60cb903

Use the spatial and temporal search and specify the period you want.

You can estimate the annual means from the monthly values above, using a data analysis tool like CDO.

Do you have the monthly data for each year?

You can download the monthly mean netcdf files from this link:

http://mirador.gsfc.nasa.gov/cgi-bin/mirador/presentNavigation.pl?tree=project&&dataGroup=Gridded&project=TRMM&dataset=3B43:%20Monthly%200.25%20x%200.25%20degree%20merged%20TRMM%20and%20other%20sources%20estimates&version=007&CGISESSID=851b1fb9546fc5ea49699ee2c60cb903

Use the spatial and temporal search and specify the period you want.

You can estimate the annual means from the monthly values above, using a data analysis tool like CDO.

Hi Vinay, The Climate Data Library from the International Research Institute for Climate and Society

http://iridl.ldeo.columbia.edu/

Affords the ability to do both of those calculation and much more.

Using the Expert mode coding interface, I have used Dakar, Senegal as an example below.

The following code sequence is what should be input into the Expert Mode box. Line by line, what does this mean?

1. The data set being used

2. The temporal range of the calculation

3. Latitude

4. longitude

5. This stipulates that only months that have at least 92% of non-missing data be calculated into the monthly average

SOURCES .NASA .GES-DAAC .TRMM_L3 .TRMM_3B42 .v7 .daily .precipitation

T (1 Jan 2000) (31 Jan 2010) RANGE

X -17.5 VALUE

Y 14.7 VALUE

T 0.92 monthlyAverage

After pressing OK, clicking on "Tables" and then "Columnar Tables" will yield the following data.

http://iridl.ldeo.columbia.edu/SOURCES/.NASA/.GES-DAAC/.TRMM_L3/.TRMM_3B42/.v7/.daily/.precipitation/T/%281%20Jan%202000%29%2831%20Jan%202010%29RANGE/X/-17.5/VALUE/Y/14.7/VALUE/T/0.92/monthlyAverage/T+exch+table-+text+text+skipanyNaN+-table+.html

Using the data library, additional calculations can be completed, or the table data can be exported in various forms for external manipulation.

you also can get TRMM 3B42 through Giovanni-interctive Visualization and analysis from the folloing link:

http://gdata1.sci.gsfc.nasa.gov/daac-bin/G3/gui.cgi?instance_id=TRMM_3B42RT

I prefer Giovanni interface, as Xiquan suggests.

But make sure that you choose archive data (not Real Time Data) which is validated by ground rain gauges over monthly readings.

You need to divide the area of interest to grids similar to TRMM pixels. Then get archived data sets for daily or weekly rainfall for the region and generate your own monthly means. A few station data also can be collected on daily rainfall and avalidation of TRMM monthly means can be carried out to know the error bars. I did a similar exercise for aerosol data from MODIS satellite.

Best regards.

Doing the computation is fine of course, but how to judge the credibility of it is another issue. The spatiotemporal representation is an issue as local rain gauges have a different daily rain PDF than a PDF obtained from daily sums from a few TRMM passes. So, the rain rate PDF may be correct over a typical TRMM pixel, but not validate well with a local rain gauge, e.g., in areas with orography. The daily temporal error due to non-uniform TRMM sampling, as compared to the uniform sampling by a rain  gauge will be substantial as well, e.g., daily afternoon rain could be missed by TRMM systematically. An attempt of error analysis is given in: Alemohammad et al., Hydrol. Earth Syst. Sci. Discuss., 12, 2527–2559, 2015

www.hydrol-earth-syst-sci-discuss.net/12/2527/2015/

doi:10.5194/hessd-12-2527-2015