Search

Gridded precipitation (PRP) data have been largely used in diagnostic studies on the climate variability in several time scales, as well as to validate model results. The three most used gauge-based PRP datasets are from the Global Precipitation Climatology Centre (GPCC), University of Delaware (UDEL), and Climate Research Unit (CRU). This paper evaluates the performance of these datasets in reproducing spatiotemporal PRP climatological features over the entire South America (SA) for the 1901–2015 period, aiming to identify the differences and similarities among the datasets as well as time intervals and areas with potential uncertainties involved with these datasets. Comparisons of the PRP annual means and variances between the 1901–2015 period and the non-overlapping 30-year subperiods of 1901–1930, 1931–1960, 1961–1990, and the 25-year subperiod of 1991–2015 for each dataset show varying means of the annual PRP over SA depending on the subperiod and dataset. Consistent patterns among datasets are found in most of southeastern SA and southeastern Brazil, where they evolved gradually from less to more rainy conditions from 1901–1930 to the 1991–2015 subperiod. All three datasets present limitations and uncertainties in regions with poor coverage of gauge stations, where the differences among datasets are more pronounced. In particular, the GPCC presents reduced PRP variability in an extensive area west of 50° W and north of 20° S during the 1901–1930 subperiod. In monthly time scale, PRP time series in two areas show differences among the datasets for periods before 1941, which are likely due to spurious or missing data: central Bolivia (CBO), and central Brazil (CBR). The GPCC has less monthly variability before 1940 than the other two datasets in these two areas, and UDEL presents reduced monthly variability before 1940 and spurious monthly values from May to September of the years from 1929 to 1941 in CBO. Thus, studies with these three datasets might lead to different results depending on the study domain and period of analysis, in particular for those including years before 1941. The results here might be relevant for future diagnostic and modelling studies on climate variability from interannual to multidecadal time scales. [ABSTRACT FROM AUTHOR]

Background: Global warming and climate change are threats to the world. Warmer temperatures and changes in precipitation patterns alter water availability and increase the occurrence of extreme weather events. South America and the Andes are vulnerable regions to climate change due to inequity and the uneven distribution of resources. Climate change evaluation often relies on the use of general circulation models (GCMs). However, the spatial resolution is too coarse and does not provide a realistic climate representation at a local level. This is of particular importance in mountain areas such as the Andes range, where the heterogeneous topography requires a finer spatial resolution to represent the local physical processes. To this end, statistical and/or dynamical downscaling methods are required. Several approaches and applications of downscaling procedures have been carried out in the countries of this region, with different purposes and performances. However, the main objective is to improve the representation of meteorological variables such as precipitation and temperature. A systematic review of these downscaling applications will identify the performance of the methods applied in the Andes region for the downscaling of precipitation and temperature. In addition, the meta-analysis could detect factors influencing the performance. The overall goal is to highlight promising methods in terms of fitness for use and identify knowledge gaps in the region. Methods: The review will search and examine published and grey literature on downscaling applications of temperature and precipitation in the Andes region. Predetermined criteria for eligibility will allow the screening of the evidence. Then, the method used in each application will be coded and mapped according to the country, purpose, variable, and type of downscaling. At the same time, quantitative and qualitative data will be extracted. The performance metrics are particularly interesting for this review. A meta-analysis will be conducted for those studies with comparable metrics. A narrative synthesis, maps and heatmaps will show the results. Tables, funnel plots, and meta-regressions will present the meta-analysis. Throughout the review, a critical appraisal step will categorize the validity of the evidence. [ABSTRACT FROM AUTHOR]