Publication details for Dr Isabella BovoloBovolo, C. Isabella, Pereira, Ryan, Parkin, Geoff, Kilsby, Chris & Wagner, Thomas (2012). Fine-scale regional climate patterns in the Guianas, tropical South America, based on observations and reanalysis data. International Journal of Climatology 32(11): 1665-1689.
- Publication type: Journal Article
- ISSN/ISBN: 0899-8418, 1097-0088
- DOI: 10.1002/joc.2387
- Further publication details on publisher web site
Author(s) from Durham
The dense, intact rainforests of the Guianas (northern Amazonia) are important for the regulation of the local and regional climate, but preclude easy access so a large data gap exists. The rainforest–savannah boundary may also be particularly susceptible to increasing pressures from ecosystem exploitation and climate change. It is important therefore to establish a baseline climate regime so that the impacts of any future change can be determined. Currently this area is not covered by the publicly available climate data sets, so interpolated data sets are not an accurate representation of the climate of this area, emphasizing the need for a more detailed analysis in this region. Here, we collate the available data sets and use these to derive maps of observed precipitation and temperature across the region. To overcome the limitations in the inadequate observational data sets, we also test and use the ECMWF ERA‐40 reanalysis data set for the period 1958–2001 at a ∼125 km2 (1.125°) resolution. Mean differences (biases) and annual average spatial correlations are examined between modelled and observed time series comparing the seasonal cycles and the yearly, monthly and monthly anomaly time series. The results show that with the exception of sub‐grid resolution mountain environments, reanalysis data for the Guianas provide a consistent and relatively accurate spatial distribution of temperature. Precipitation is modelled less accurately, with best results for the average timings, length and severity of the dry periods. This study provides consistent and quantitative details on the spatial variations in seasonality, particularly in areas lacking observations, thereby advancing beyond previous studies. Precipitation is highly variable in the region so care must be taken when averaging modelled data over large geographical areas for comparisons with gridded data sets based on few observations. This is the first comprehensive study of the recent historical climate and its variability in this area.