Abstract: The datasets of the daily precipitation and temperature at 58 stations in the cultivating areas in Inner Mongolia for the past 50 years (1961-2010) and the yield perunit records of 38 leagues (cities) for the past 48 years (1961-2008) of wheat, 30 years (1979-2008) of maize and potato were collected, investigated and analyzed for preestimating the variations of the three major crop yields for the future 30 years (2011-2040) In this research, the authors utilized natural logarithm function to present the crop trend yield, and then separated the meteorological yield from the observed yield series Furthermore, results of the multiple linear regressions indicated that the growing season precipitation and temperature made significant impacts on the three major crops in most leagues (cities) of Inner Mongolia In addition, regional climate simulating outputs of the PRECIS under the SRES for A1B, A2 and B2 emission scenarios, and the CMIP5 multi model dataset were employed to simulate climate conditions for the future 30 years The meteorological yield, simulated temperature and precipitation for all leagues (cities) were composed into the multiple linear regression models for preestimating the coming 30 years meteorological yield of the three crops The regression results indicated that all meteorological yields perform an increasing trend, though in several years display decreased yield The preestimated quantity of meteorological yield perunit for wheat, maize and potato would respectively increase about 1790, 516 and 507kg/ha during the future 30 years The yield displayed an uptrend in all leagues(cities) after the natural logarithm trend yield to be added into the meteorological yield series, and the yield predicted would increase by 12214, 21211 and 10080kg/ha for wheat, maize and potato wheat Results of this study can be referenced as important scientific information for facing future climate change and making developing policy for future crop production

Key words: Logarithm trend yield, Meteorological yield, Simulated climate conditions, Yield preestimation, Meteorological risks