The high risk of agricultural activities demands the adoption of risk-assessment methods based on the integration of climatic and pedologic parameters, thus allowing the establishment of state policies concerning crop insurance. Based on the Brazilian Land Evaluation System, a methodology was drawn to elaborate a pedoclimatic zoning of the lands in the State of Goiás and Federal District for the planting of rainfed rice (Oriza sativa L.) (110 days-cycle). Fertility deficiency was assessed in relation to the sum of bases, base saturation and, especially, the CEC of the soils. The data were obtained from RADAMBRASIL's soil survey reports (Goiânia, Goiás and Brasília folios, in the scale of 1:1,000,000). Water deficiency was assessed by means of a study on the climatic risks for the crop. Other sources of data were: daily rainfall information from 121 pluviometric stations, ETp data calculated for 21 weather stations, available water capacity in the main types of soil, as well as the crop coefficients and growth stages, and the effective depth of rice roots. From the information above, the SARRA model simulated the water balance for four different planting dates. Of the lands, 52.8 and 11.1%, respectively, were rated as having medium and low limitations due to fertility deficiency. The aptitude of lands is considered as a dynamic concept, since the degree of water deficiency changes with the date of planting. In the case of rice planted in October, the aptitude was considered good , regular, or very restricted in 15.2, 48.3, and 0.4% of the lands, respectively. In contrast, for rice planted in January, the figures are 3.6, 26.2, and 34.0% for the same categories. The results should be seen as yardsticks for assessing the suitability of lands in Goiás and the Federal District for the planting of rainfed rice. They may be used as guidelines for agrarian planning and policy in the region.
land suitability; rainfed rice; water balance; GIS; cerrados