Testing gridded land precipitation data and precipitation and runoff reanalyses (1982–2010) between 45° S and 45° N with normalised difference vegetation index data

Abstract. The realistic simulation of key components of the land-surface hydrological cycle – precipitation, runoff, evaporation and transpiration, in general circulation models of the atmosphere – is crucial to assess adverse weather impacts on environment and society. Here, gridded precipitation data from observations and precipitation and runoff fields from reanalyses were tested with satellite derived global vegetation index data for 1982–2010 and latitudes between 45° S and 45° N. Data were obtained from the Climate Research Unit (CRU), the Global Precipitation Climatology Project (GPCP) and Tropical Rainfall Monitoring Mission (TRMM; analysed for 1998–2010 only) and precipitation and runoff reanalyses were obtained from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR), the European Centre for Medium-Range Weather Forecasts (ECMWF) and the NASA Global Modelling and Assimilation Office (GMAO). Annual land-surface precipitation was converted to annual potential vegetation net primary productivity (NPP) and was compared to mean annual normalised difference vegetation index (NDVI) data measured by the Advanced Very High Resolution Radiometer (AVHRR; 1982–1999) and Moderate Resolution Imaging Spectroradiometer (MODIS; 2001–2010). The effect of spatial resolution on the agreement between NPP and NDVI was investigated as well. The CRU and TRMM derived NPP agreed most closely with the NDVI data. The GPCP data showed weaker spatial agreement, largely because of their lower spatial resolution, but similar temporal agreement. MERRA Land and ERA Interim precipitation reanalyses showed similar spatial agreement to the GPCP data and good temporal agreement in semi-arid regions of the Americas, Asia, Australia and southern Africa. The NCEP/NCAR reanalysis showed the lowest spatial agreement, which could only in part be explained by its lower spatial resolution. No reanalysis showed realistic interannual precipitation variations for northern tropical Africa. Inclusion of runoff in the NPP prediction resulted only in marginally better agreement for the MERRA Land reanalysis and slightly worse agreement for the NCEP/NCAR and ERA Interim reanalyses.