The climate of the Cerrado region is controlled by the inter-tropical convergence zone (ITCZ) and the South Atlantic convergence zone (SACZ). The South Atlantic monsoon system occurs east of the Andes, with precipitation maxima in the convergence zones in summer. The monsoon season is influenced by the El Niño Southern Oscillation (ENSO) (IPCC 2013; WG1).
Past and future climate trends for the Cerrado are available through trend analysis of the SREX region VII covering the upper-central part of South-America (SREX: Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation, SREX region numbering according to IPCC-WG1 report, 2013). Furthermore, projections with regional models are available through the studies conducted by Marengo et al. (2009, 2010).

Past trends in the Cerrado
Past climate trends across central and northern Brazil show a general temperature increase at a rate of 0.08 to 0.2 °C per decade; opposite trends regarding precipitation range between -50% and +20% of the annual rate over the last 50-100 years (IPCC 2007a, WG1). In the southern Amazon, the mean temperature increased by 0.08 °C per decade during 1901-2001. During 1948-1999, the precipitation trends exhibited changes between -23% and +18 % (Marengo 2004). The IPCC (2013, WG1) reports an increased occurrence of extreme rainfall events, but the lack of long-term records make robust trend analysis impossible (Liebmann and Mechoso 2011). However, Assad et al. (1993) analysed 100 rainfall stations over a minimum of 20 years (1940-1970) across the Cerrado to identify a return period of 10-30 consecutive dry days during the rainy season. The authors found that for a 5 year return period, 25 days without rainfall occurred in the states of Piauí, South Maranhão, and the northern Minas Gerais. In contrast, Mato Grosso, Goiás, and the triangle of Mineiro showed shorter dry spell lengths of 10 days. Carvalho et al. (2013) repeated the analysis with a time series of 70 years and provided evidence for an intensification in the dry spell length. For the Midwest region, dry spells of 5-25 days were observed with a one-year return period and 40-50 days without rainfall were observed for return periods of 4 to 5 years. In addition, in the last 30 years, El Niño (ENSO) has occurred more often in the southern Amazon Basin and has resulted in an increased susceptibility of wild fire occurrence (Nepstad et al., 2004). However, La Niña has also increased due to climate change and has caused natural wildfires in the wet season via lightning strikes, which doubled between 2005 and 2008 (Pivello 2011).

Future projections in the Cerrado
Future predictions of temperature trends for the large SREX regions comprising the Cerrado, as derived by a set of 42 global models (CMIP5), estimate a temperature increase of 0.5 -1.8 °C by 2035 (2016-2035) for the new RCP4.5 scenario. In a regional analysis (A1B emission scenario, Eta-CPTEC) of the Paraná Basin, which covers parts of the central Cerrado, Marengo et al. (2012) similarly estimated a temperature increase of 1.8 °C between 2011 and 2040. Global circulation models (GCM) project a warming of up to 3.3 °C in the region in both seasons by 2065 (2046-2065) and a warming of up to 4 °C by 2100 (IPCC, 2013), which is also projected for the Cerrado based on regional models (Marengo et al., 2010); therefore, the Cerrado is predicted to warm faster than the global average (Marengo et al., 2011). Using deforestation scenarios from Soares-Filho et al. (2006), Costa and Pires (2010) demonstrated that the concurrent deforestation in the Cerrado directly affects the future climate of the Amazon rainforest. Deforestation will reduce rainfall in April and September-November and prolong the dry season from five to six months in the Cerrado.
In Brazil, droughts are very likely to substantially increase (IPCC 2013, WG1), particularly in the NE-Cerrado (Marengo et al., 2010). GCM precipitation projections for NE-Brazil are variable, ranging from -19% to +17% for 2016-2035; however, in the long-term, it is very likely that less rainfall will occur in the eastern Cerrado in the dry season (high confidence for 2081-2100, IPCC 2013). In the Paraná River Basin, Marengo et al. (2012) projected a slight precipitation decrease of 2.1% between 2011 and 2040 for both seasons using regional models. The decrease is projected to be as high as 12% in winter; the summer precipitation will remain approximately the same (-0.7%). In the simulations for 2071-2100 under the A2 scenario, Marengo et al. (2010) found that rainfall will decrease in NE-Brazil; in the central Cerrado area, one model shows increases while others project reductions. However, heavy rainfall events greater than 10 mm are very likely to increase in the Cerrado (Marengo et al., 2009, Marengo et al., 2010, IPCC 2013).

Model uncertainty
Because of inconsistencies between the different climate models, both the IPCC reports (2007, 2013) and Marengo et al. (2010) conclude that current GCMs and RCMs are not able to project changes in future rainfall regimes at regional scales for the west-central region in Brazil. Most of the climate change prediction models neglected to include future land use dynamics, which potentially have a significant effect on the future micro-climate (Costa and Pires 2010). For example Loarie et al., (2011) reported for the Cerrado that the conversion of natural vegetation to crop/pasture had led to a warming of in average 1.55 °C whereas the conversion of crop/pasture to sugarcane caused a cooling of 0.9°C. Projections of the future temperature and rainfall regimes are crucial to develop management plans for future agricultural practice; however, all in all, current climatic trend projections for the region appear to be highly insufficient to do so.