Rainfall is important in the hydrological cycle and to the lives and welfare of humans. In addition to being a life-giving resource, rainfall processes also plays a crucial role in the dynamics of the global atmospheric circulation. One unresolved aspect of this problem is the climate trend found in some products but not others, as well as the far larger inter-seasonal variability of observed precipitation relative to general circulation and climate models. This has important consequences for the atmospheric energy budget and thus needs to be reconciled. The TRMM satellite was launched on Nov. 27, 1997 in order to begin quantifying the rainfall amounts, rainfall structure and latent heating in the tropics. While the overall magnitude and location of precipitation appears to be fairly well constrained by this satellite, variability and trends from different sensor are still showing discrepancies. These discrepancies are found to be the result not of algorithm errors, but of a number of parameters in the algorithms which must be assumed constant but in reality vary with both space and time. It is this variability of the rain structures which in turn causes precipitation time series to differ among the different sensors used to retrieve rainfall. This talk will illustrate the phenomenon and show how ground based observations can be used to separate real changes from those that might be due to changes in the rainfall structure.