Cycles and trends are considered structural information in historical data which, if detectable and quantifiable, can be used in simulating future climates. Three data-processing models were applied to climatic records to test detectability of such parameters of climatic change. One was the conventional ARIMA model. The other two were developed and tested for this study. The first model detected cycles of varying wavelength in data. It tested successfully with monthly data with known seasonal cycle. The second model detected persistence of high and low values during cycling, and included a time-trend component. This model tested successfully against purposefully designed synthetic data. The three models were applied to two rainfall and two temperature records in Georgia. No model detected any significant structure in these climatic records of 52-104 years. It is concluded that short-term climatic simulation for risk and uncertainty analysis in agricultural planning in this region need not presently include any time-shifting parameters. However, since slight climatic change may not be detected in single-site analysis, future research is suggested for multisite analyses. This publication will be useful to anyone concerned with examining climatic records for natural and human-induced cycles and trends.