在充分搜集河北省退耕还林前后粮食生产和农民人均纯收入的动态变化有关情况的基础上,应用VARMA预测模型,预测了2006年到2015年河北省粮食生产情况,并用2006年和2007年的各项实际数据与预测数据进行对比核对。预测结果表明,各项数据指标的实际检验误差都在误差范围之内,说明所建立的预测模型具有较高的精确度,为今后制定和完善退耕还林后续政策提供了可靠依据。 Based on the full analysis of the dynamic changes of grain production and per capita net income of farmers before and after the conversion of cropland to forest in China, the VARMA prediction model was used to predict the grain production in Hebei Province from 2006 to 2015, The actual data and forecast data for comparison check. The prediction results show that the actual test error of each data index is within the error range, indicating that the established prediction model has higher accuracy and provides a reliable basis for formulating and perfecting the follow-up policy of returning farmland to forest in the future.