|Coverage||Global (90N-90S, 0-360E)|
|Time Period||January 1979 - Present|
|Resolution||2.5 x 2.5 degree, monthly|
Several climate rainfall products, which merge rainfall estimates from a variety of satellite and ground-based sources are currently available going back to January of 1979 and continuing through the present. While the combination of data from these multiple sensors/sources provides long-term climate rainfall datasets suitable for climate studies, there are a number of issues affecting their use for some climate applications. These include discontinuities in the component data sets, differences in the calibration methods, and the methodology used to weight the individual rain estimates.
The current version is version 0802.
The CMAP dataset merges satellite and rain gauge data from a number of satellite sources and rain gauge sources. This version of the CMAP dataset also uses precipitation from Numerical Weather Prediction (NWP) models. The NWP data is used primarily to fill in gaps at high latitudes. For users comparing with or validating NWP models, there is a version of CMAP using observations only for its rain estimates. Details on the component datasets as well as the method used to merge these data are provided by Xie and Arkin [1996, 1997].
This dataset is available through the
In an attempt to explain discrepancies in interannual variability of tropical mean rainfall between PR (2A25) and TMI (2A12) estimates, Robertson et al.  compared monthly mean values of the PR path integrated attenuation (PIA) with surface rain rates. They found that the variability in PIA closely matched variability in the TMI rain estimates, suggesting that uncertainties in the assumed drop size distribution and the associated attenuation/reflectivity/rainfall relationships inherent in single-frequency radar methods is a serious issue for climate studies.