中国岩溶
中國巖溶
중국암용
CARSOLOGICA SINICA
2009年
4期
426-431
,共6页
灰色预测%GM(1,1)模型%Verhulst模型%喀斯特%耕地面积变化
灰色預測%GM(1,1)模型%Verhulst模型%喀斯特%耕地麵積變化
회색예측%GM(1,1)모형%Verhulst모형%객사특%경지면적변화
gray forecast%GM (1,1) model%Verhulst model%karst%forecast of farmland
毕节生态试验区是典型的喀斯特地区,人口密度大,生态环境脆弱,水土流失严重,耕地保护尤为重要.本文以毕节地区1998-2006年常用耕地面积统计数据为基础,分别利用灰色GM(1,1)模型和Verhulst模型对该喀斯特地区常用耕地面积变化作预测.预测结果表明,GM(1,1)模型和Verhulst模型都揭示了毕节地区常用耕地面积在未来几年呈逐年递减的趋势.但是通过模拟精度验证,研究区耕地面积原始数据变化曲线呈S型,就灰色GM(1,1)模型与Verhulst模型模拟预测结果相比较,对于原始数据的模拟,GM(1,1)模型预测精度较高,而Verhulst模型更适合研究区未来几年耕地面积变化的预测.该研究结果可为区域合理利用土地资源、编制土地利用规划和耕地保护提供依据.
畢節生態試驗區是典型的喀斯特地區,人口密度大,生態環境脆弱,水土流失嚴重,耕地保護尤為重要.本文以畢節地區1998-2006年常用耕地麵積統計數據為基礎,分彆利用灰色GM(1,1)模型和Verhulst模型對該喀斯特地區常用耕地麵積變化作預測.預測結果錶明,GM(1,1)模型和Verhulst模型都揭示瞭畢節地區常用耕地麵積在未來幾年呈逐年遞減的趨勢.但是通過模擬精度驗證,研究區耕地麵積原始數據變化麯線呈S型,就灰色GM(1,1)模型與Verhulst模型模擬預測結果相比較,對于原始數據的模擬,GM(1,1)模型預測精度較高,而Verhulst模型更適閤研究區未來幾年耕地麵積變化的預測.該研究結果可為區域閤理利用土地資源、編製土地利用規劃和耕地保護提供依據.
필절생태시험구시전형적객사특지구,인구밀도대,생태배경취약,수토류실엄중,경지보호우위중요.본문이필절지구1998-2006년상용경지면적통계수거위기출,분별이용회색GM(1,1)모형화Verhulst모형대해객사특지구상용경지면적변화작예측.예측결과표명,GM(1,1)모형화Verhulst모형도게시료필절지구상용경지면적재미래궤년정축년체감적추세.단시통과모의정도험증,연구구경지면적원시수거변화곡선정S형,취회색GM(1,1)모형여Verhulst모형모의예측결과상비교,대우원시수거적모의,GM(1,1)모형예측정도교고,이Verhulst모형경괄합연구구미래궤년경지면적변화적예측.해연구결과가위구역합리이용토지자원、편제토지이용규화화경지보호제공의거.
As a typical karst area with a dense population and fragile ecological environment as well as serious soil erosion in Bijie ecological experimental area, the farmland protection is particularly important. Based on the statistic data of farmland from 1998 to 2006 in Bijie Region, this paper forecasts the changes of farmland in karst region with GM (1,1) model and the Verhulst model respectively. The forecast results show that both GM (1,1) model and the Verhulst model reveal the cultivated land in Bijie Region decreases gradually in the next years. However, by the NC verification of the model, the change of the original data of the cultivated land in the study area shows an S curve. To the imitation of the original data, the prediction results of the grey GM (1,1) are more accuracy than the Verhulst model, while the Verhulst is more suitable for the forecast of the changes in the coming years of the cultivated land in the study area. It can provide the basis for the rational use of land resources, the preparation of land-use planning and the protection of farmland.