Abstract:
2.2 数据分析方法本文采用统计分析和空间分析方法,分析等别平衡转换方法的特点,并验证其转换精度。1)统计分析方法:本文以西部区12省农用地调查样点数据为分析样本,采用EXCEL和SPSS软件,对样点数据进行线性分析与显著性检验,并分析转换前后省级等指数与国家级等指数的关系。2)空间分析方法:采用Arcgis的空间分析方法,对成果进行转化,实现在空间上对国家级成果进行表达与分析。3 结果与分析3.1 等别平衡转换3.1.1 建立利用等指数与标准粮产量的线性回归模型运用SPSS软件,以西部区12个省样点为分析样本,以实际标准粮产量为因变量y,利用等指数为自变量x,进行线性回归分析,建立了线性回归模型(表2),并对得到的线性回归模型进行检验。表2 基于线性回归模型的西部各省(市、区)等别平衡转换表Table 2 Linear equations of the transferring arable land grade省份Provinces样点数Number of sample省级等指数与标准粮产量回归方程Regressive equations between standard crop yields and grading index at province levelR2省级等指数向国家级等指数平衡转换规则Transferring rules between grading index from province to national level省等指数范围Range of grading index at provincal level国家级等指数范围Range of grading index at national level国家级等别范围Range of grade at national level内蒙古454y=13.40x1+23430.69x2=1.79 x1 +3122~91025~19776~15广 西547y=4.33x1+ 46260.88x2=0.58 x1 +617352~2943819~23164~11重 庆959y=5.5 x1+ 13500.65x2=0.75 x1 +180575~2064608~17187~12四 川316y=5.37 x1+ 32790.91x2=0.72 x1 +437103~2709510~23784~13贵 州625y = 5.24 x1+ 7870.88x2=0.70 x1 +105238~2273271~16937~14云 南1749y=4.32 x1+ 37260.80x2=0.56 x1 +54050~3090568~22704~13西 藏279y= 5.81 x1+ 6730.68x2=0.77 x1 +90347~1369359~115010~14陕 西1712y = 6.65 x1+9420.86x2=0.89 x1 +126149~2226257~20965~14甘 肃2128y = 6.51 x1+ 15550.85x2=0.87 x1 +20720~1773225~17477~14青 海364y=4.05 x1+ 15830.66x2=0.54 x1 +211147~1263290~89211~14宁 夏1510y=10.38 x1+22840.90x2=1.38 x1+30417~1175387~19316~14新 疆235y=4.46 x1+ 18240.84x2=0.59 x1+24311.4~2904250~19706~14注:y为标准粮产量standard crop yields,x1为省级利用等指数grading index at provincial level,x2为国家级利用等指数grading index at national level.由表2可知,各省(市、区)的省级利用等指数与标准粮产量之间的线性回归方程有较大差异,这是由于各省(市、区)分等因素指标体系及指定作物、产量比系数、土地利用系数等相关分等参数的差异31,造成各省(市、区)利用等指数的内涵不统一所致26。因此,需要依据回归模型对各省级利用等指数进行平衡转换,以建立各省(市、区)利用等指数与耕地生产能力关系的一致性。从各线性回归方程的相关系数来看,西部12省(市、区)的样点数均大于100,根据相关系数检验表,当自由度为100时,置信水平为0.01时,相关系数R2=0.0645。西部各省(市、区)省级利用等指数与标准粮产量之间线性回归方程的R2值均大于0.0645,表明各省(市、区)的省级利用等指数与标准粮产量相关,所建立的线性回归模型可以做为等别平衡转换的依据。3.1.2 确定省级利用等指数向国家级利用等指数转换规则并进行转换根据确定的国家级利用等指数与实际标准粮产量的相关方程y=7.5x2,以及建立的各省级利用等指数与实际标准粮产量线性回归模型,初步分析得到省级利用等指数向国家级利用等指数平衡转换规则(表2)。应用分析得到的各省级利用等指数向国家级利用等指数平衡转换的规则,将各省级利用等指数转换为国家级利用等指数,按照国家确定的200分等间距划分国家级利用等别(表2)。3.2 等别转换结果精度分析3.2.1 等别转换结果模型验证在全国农用地分等工作过程中,选择最高利用水平的典型耕地地块,建立了国家级、省级、县级3级标准样地体系,全国共设立了国家级标准样地约500块,省级标准样地约5 000块,县级标准样地约50 000块。一般情况下,标准样地在区域内所处位置的气候、地形地貌、土壤、灌溉与排水等条件的综合特征最优,具有区域代表性,而且标准样地是独立于上文建模样点之外的地块。因此,可采用实地调查得到的标准样地属性数据对转换结果进行模型检验,用于检验转换结果的精度。本文采用西部区省级标准样地数据,对西部区重点省份的等别平衡转换结果进行模型验证。参与验证的省份包括内蒙古、四川、贵州、云南、陕西、甘肃、新疆7个省(自治区),共选取省级标准样地1 088块(表3),各样地的属性包括自然条件、实际标准粮产量和农用地分等结果数据等。验证方法如下:1)将各标准样地的实际标准粮产量定义为y1;2)依据表2中的各省级等指数与标准粮产量的回归方程,计算得出各标准样地样点的模拟标准粮产量y2;3)将各样地模拟标准粮产量y2做为自变量,将各样地实际标准粮产量y1做为因变量,进行线性回归分析,建立各重点省份标准样地的分析模拟标准粮产量与实际标准粮产量之间的线性回归方程(表3);4)计算样地模拟标准粮产量的平均绝对误差和平均相对误差。通过表3可以看出,各省模拟标准粮产量与实际标准粮产量之间相关性较高,除个别省份存在较大误差外,总体上误差较小,验证了农用地平衡转换方法的准确性。表3 标准样地样点的模拟标准粮产量与实际标准粮产量相关关系Table 3 Correlation between actual standard crop yields and stimulant standard crop yields of standard-plot省份Provinces所选标准样地数量/块Number of standard-plot模拟标准粮产量与实际标准粮产量回归方程Regressive equations between actual standard crop yields and stimulant standard crop yieldsR2平均绝对误差Mean absolute error平均相对误差Mean relative error/%内蒙古95y1=1.00y2-540.687519.25四 川261y1=0.94y2+7770.95-57-0.03贵 州119y1=0.99y2-8780.6596812.70云 南37y1=0.96y2+31840.83-275320.62陕 西300y1=0.99y2+1390.849.264.50甘 肃231y1=0.94y2-190.694289.24新 疆45y1=0.88y2+8700.64-180.84合 计1088y1=1.02y2-2960.8697.95.60注:y1为实际标准粮产量actual standard crop yields;y2为模拟标准粮产量stimulant standard crop yields.3.2.2 省级成果与国家级成果的关系1)省级利用等指数及利用等别在省级范围内可以比较。通过以上建立的西部各省(市、区)省级利用等指数与标准粮产量线性回归方程,可以看出各省(市、区)省级利用等指数与标准粮产量呈现高度的相关性,因此,省级利用等指数的大小体现了省内耕地生产能力的高低,保障了省级利用等指数及利用等别在省内的统一可以比较。西部各省(市、区)省级利用等指数与标准粮产量线性回归方程存在较大差异,说明省际间省级利用等指数与标准粮产量的换算关系不一致,即不同省份利用等指数的内涵不统一,因此省级利用等指数及利用等别在省际间不具有可比性。2)国家级利用等指数及利用等别继承了省级利用等指数及利用等别的高低顺序。从利用等指数的转换过程来看,平衡转换的方法是以标准粮产量为媒介,建立了省级利用等指数与国家级利用等指数的线性转换关系,使得国家级利用等指数与省级利用等指数存在一一对应的关系,因此,国家级利用等指数及利用等别继承了原省级利用等指数及利用等别的高低顺序。3)国家级利用等指数及利用等别在省际范围内可以比较。依据平衡转换方法,得到的国家级成果,使省与省之间的农用地利用等指数具有相同的标准粮内涵,实现了国家级利用等指数及利用等别在省际间的统一可以比较。3.2.3 西部12省国家级利用等别转换结果通过上文分析,省级利用等指数转换为国家级利用等指数后,在省际间具有可比性,则依据国家级利用等指数采用200分等间距划分的国家级利用等别在省际间宜具有可比性。经过等别平衡转换,将西部区12省的耕地划分为12个国家级利用等别,等别范围为4~15等,4等地生产能力最高,15等地生产能力最低。西部12省国家级利用等别空间分布情况如图2。可以看出,西部区耕地分布较少、破碎化程度高,耕地质量总体水平较低,耕地等别以10~15等为主,总体上南部地区的耕地利用等别优于北部地区,平原区的耕地利用等别优于山地区和高原区。西部各省(市、区)国家级利用等别分布情况有一定差异,广西、四川、重庆等省份耕地利用等别较高,等别集中在6~10等之间;内蒙、宁夏、甘肃、陕西等省份的耕地利用等别较低,其等别多集中在10等以后。可见,转换后的国家级利用等别分布特征总体上与区域资源本底一致,表明国家级利用等别不仅较好的继承了省级利用等别的高底序列及其在生产能力方面的一致性,同时,转换后的农用地分等成果与资源本底之间具有较好的空间相关性。图2 西部区12省耕地国家级利用等别空间分布图Fig.2 Spatial distribution of arable land grade in the western of China4 结论与讨论1)西部12个省的省级农用地分等指数与标准量产量高度相关,依据标准量产量,实现了省级农用地等别向国家级等别的转化,并保障了等别的统一标准粮内涵和统一可以比较。2)转化后的西部12省的农用地等别被转化为12个国家级利用等别,等别范围为4~15等,耕地等别总体偏低,等别分布特征与区域资源本底总体一致。3)对重点省份的转换模型进行验证,结果表明转换模型精度较高,等别平衡转转方法有效的实现了全国耕地等别的统一可比。4)平衡转换方法是一种行之有效的农用地成果汇总方法,其有效地解决了省级成果之间等别内涵不统一,等别指数不可比的汇总难题,建立起国家级统一可以比较的成果体系,实现了国家级成果和省级成果的一一对应关系,从而保障了国家级成果的统一可以比较,保证了省级成果和国家级成果差异化管理的需要。5)省级成果在省级范围内比较,可以满足在省级范围内,建立适合区域特点的耕地质量利用保护与质量提升对策;国家级成果在省际之间可以比较,为在国家宏观层面实施耕地质量利用与保护对策,提供了科学依据。两级成果的独立性,为实施差异化的耕地质量管理对策提供了技术支撑。标准粮产量是决定农用地分等指数是否转化成功的重要参数,因此,如何科学准确计算标准粮产量是需要进一步深入研究的问题。本文统一采用建立线性回归模型的等别平衡转换方法完成了西部区省级等别向国家级等别的转换,并对转换结果进行了分析。然而,到底如何更加科学简单的进行转化,提高成果转化精度,也是未来农用地分等要考虑的问题。参 考 文 献1 GB/T 28407-2012,农用地质量分等规程s.2 Yun Wenju, Cheng Feng, Zhang Yuanyu, et al. The cultivated land quality in China: Baseline investigation and a long-term monitoring programC// Land and Water Management, Decision Tools and Practices, Proceedings of the 7th Interregional Conference on Environment and Water, Beijing, China Science and Technology Press, 2004: 550-558.3 张凤荣,郧文聚,胡存志,等. 《农用地分等规程》的几个理论问题及应用方向J. 资源科学,2005,27(2):33-38.Zhang Fengrong, Yun Wenju, Hu Cunzhi. The theory and application of the regulations of farm land gradingJ. Resources Science, 2005, 27(2): 33-38. (in Chinese with English abstract)4 王洪波. 农用地分等定级的理论与方法探讨J. 土壤,2004,36(4):371-375.Wang Hongbo. Theory and method of grading and rating of agricultural landJ. Soils, 2004, 36(4): 371-375. (in Chinese with English abstract)5 高向军,马仁会. 中国农用土地等级评价研究进展J. 农业工程学报,2002,18(1):165-168.Gang Xiangjun, Ma Renhui. Research advances of gradation and evaluation of agricultural land in ChinaJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2002, 18(1): 165-168. (in Chinese with English abstract)6 王令超,王国强,杨建波,等. 农用地分等中作物生产潜力空间插值方法研究J. 农业工程学报,2006,22(2):89-93.Wang Lingchao, Wang Guoqiang, Yang Jianbo, et al. Spatial interpolation method of crop productive potentiality for the classification of agricultural landJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2006, 22(2): 89-93. (in Chinese with English abstract)7 李如海,周生路,宋佳波,等. 农用地分等指标区与参评因素定量确定J. 土壤学报,2004,41(4):517-522.Li Ruhai, Zhou Shenglu, Song Jiabo, et al. Quantitative determination of index areas and participating factors for farmland gradationJ. Acta Pedologica Sinica, 2004, 41(4): 517-522. (in Chinese with English abstract)8 张凤荣,郧文聚,孔祥斌,等. 对《农用地分等定级规程》土地利用系数的探讨J. 中国土地科学,2002,16(1):16-19.Zhang Fengrong, Yun Wenju, Kong Xiangbin, et al. Discussion on the land use coefficient of regulations for classification and gradation on agricultural landJ. China Land Science, 2002, 16(1): 16-19. (in Chinese with English abstract)9 马仁会,李强,崔俊辉,等. 土地经济系数宏观分区计算方法比较研究J. 农业工程学报,2005,21(增刊):159-163.Ma Renhui, Li Qiang, Cui Junhui, et al. Comparative study of the macroscopical zoning method for calculating land economic coefficientJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2005, 21(Suppl): 159-163. (in Chinese with English abstract)10 郧文聚.农用地分等及其应用研究D. 北京:中国农业大学,2005.Yun Wenju. Agricultural Land Classification and Its ApplicationD. Beijing:China Agricultural University, 2005. (in Chinese with English abstract)11 邱维理,王娟,郭文祥,等. 农用地质量是由哪些要素构成的?:农用地分等规程解读J. 中国发展,2012,12(4):54-57.Qiu Weili, Wang Juan, Guo Wenxiang, et al. Elements for the quality of agricultural land:Reading of the procedure for agricultural land gradingJ. China Development, 2012, 12(4): 54-57. (in Chinese with English abstract)12 司振中,李貌,邱维理,等. 中国耕地资源的区域差异与保护问题J. 自然资源学报,2010,25(5):713-721.Si Zhenzhong, Li Mao, Qiu Weili, et al. Regional difference analysis and preservation of cultivated land resources in ChinaJ. Journal of Natural Resources, 2010, 25(5): 713-721. (in Chinese with English abstract)13 路婕,吴克宁,郧文聚,等. 全国统一的耕地占补平衡按等级折算系数研究J. 资源与产业,2012,14(4):128-134.Lu Jie, Wu Kening, Yun Wenju, et al. National unified conversion coefficient by hierarchy on farmland requisition-compensation balanceJ. Resources and Industries, 2012, 14(4): 128-134. (in Chinese with English abstract)14 孔祥斌,张青璞. 中国西部区耕地等别空间分布特征J. 农业工程学报,2012,28(22):1-7.Kong Xiangbin, Zhang Qingpu. Spatial distribution characteristics of arable land grade in Western Ce conversion from provincial to national level; 2) the accuracy of the arable land grade at provincial level was the@principal result of this method; 3) the sci来幮汴ゥ獦杩晣@遣塡卬」筵la聴坩副筮嘠絯学簠顴乨酥瘠彳呴andar兤丠嵣穲孯扰@yiel、s of 。rabャe land was critical factor to the accuracy of the method; 4) the accomplishments of arable land grade in the western of 12 provinces at national level were in accordance with the characteristics of distribution of productivity ッf arable land. The method of the balance tra遮敳聦er乲噩聮坧笠繢譡孳乥癤洠硯穮 the 即乴。乮噤奡坲兤爠祣イop yゥeャds of a孲祡敢ャ来映ャ獡偮、笠gr乡酤步簠散孯兵畬坤删筢癥张呵sed f兯乲崠穤孩扦ヲerenエ leveャs sオch as county's and provincial levels.n, et al. Impacts of productive ratio coefficient on grade of farmland classificationJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2009, 25(1): 237-243.(in Chinese with English abstract)18 赵丽红,郭熙,陈文波. 农用地分等中土地经济系数计算方法的探讨:以江西省玉山县为例J. 土壤,2007,39(4):652-657.Zhao Lihong, Guo Xi, Chen Wenbo.Calculation methods of land economic coefficient in farmland classification:A case study of Yushan County, Jiangxi provinceJ. Soil, 2007, 39(4): 652-657. (in Chinese with English abstract)19 王洪波,郧文聚,吴次芳,等. 农用地分等图形数据库的追溯法汇总技术J. 农业工程学报,2008,24(9):59-63.Wang Hongbo, Yun Wenju, Wu Cifang, et al. Integration technology for agricultural land grading figure database based on retrospective analysis methodJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008, 24(9): 59-63. (in Chinese with English abstract)20 王国强,王令超,田燕,等. 农用地分等成果省级汇总的基础与方法研究J. 农业工程学报,2007,23(6):271-275.Wang Guoqiang, Wang Lingchao, Tian Yan, et al. Foundation and method for gathering agricultural land classification achievements at provincial levelJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2007, 23(6): 271-275. (in Chinese with English abstract.21 罗广祥,张转,弓晓敏,等. 多属性集簇农用地分等省级汇总方法J. 农业工程学报,2008,24(增刊1):145-148.Luo Guangxiang, Zhang Zhuan, Gong Xiaomin, et al. Provincial summarizing of farmland classification by poly-attributes clusterJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008, 24(Suppl.1): 145-148. (in Chinese with English abstract)22 赵寒冰,包世泰,胡月明,等. 不同比例尺农用地分等图件数据自动汇总技术研究J. 农业工程学报,2007,23(8):155-159.Zhao Hanbing, Bao Shitai, Hu Yueming, et al. Key technology for the provincial auto-integration of the attribute of different scale farmland using classification mapsJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2007, 23(8): 155-159. (in Chinese with English abstract)23 张青璞,孔祥斌,郧文聚,等. 重庆市国家级农用地分等汇总前后等别分布规律J. 农业工程学报,2010,26(10):297-303.Zhang Qingpu, Kong Xiangbin, Yun Wenju, et al. National agricultural land grading distribution law based on conversion from provincial level to national level in ChongqingJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2010, 26(10): 297-303. (in Chinese with English abstract)24 孟超,张俊梅,许皞,等. 农用地分等国家级汇总等别转换方法研究:以河北省为例J. 中国农学通报,2008,24(3):350-354.Meng Chao, Zhang Junmei, Xu Hao et al. Study on the method for grade conversion of the farmland grading collection at national Level: A Case Study in HeBei provinceJ. Chinese Agricultural Science Bulletin, 2008, 24(3): 350-354. (in Chinese with English abstract)25 杨丽辉,吴克宁,金三宝. 农用地分等中等指数与标准粮产量回归分析模型研究:以辽宁省为例J. 安徽农业科学,2008,36(19):7952-7953,8054.Yang Lihui, Wu Kening, Jin Sanbao. Study on model of regression analysis on standard grain production and medium Index in farmland classificationJ. Journal of Anhui Agri Sci, 2008, 36(19): 7952-7953, 8054. (in Chinese with English abstract)26 饶彩霞,吴克宁,许琳,等. 农用地分等中省级成果国家级汇总中的转换问题J. 农业工程学报,2008,24(增刊1):194-197.Rao Caixia, Wu Kening, Xu Lin, et al. Transfer problems in the process of farmland classification from provincial results to national rankJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008. 24 (Suppl.1): 194 一197. (in Chinese with English abstract)27 张蕾娜,郧文聚,苏强,等. 基于农用地分等成果的产能核算研究J. 农业工程学报,2008,24(增刊1):133-136.Zhang Leina, Yun Wenju, Su Qiang, et al. Methods for productivity calculation based on agricultural land classificationJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008, 24(Suppl.1): 133-136. (in Chinese with English abstract)28 伍育鹏,郧文聚,邹如. 耕地产能核算模型的研究J. 农业工程学报,2008,24(增刊2):108-113.Wu Yupeng, Yun Wenju, Zou Ru. Model for calculation of cultivated land productivityJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008, 24(Suppl.2): 108-113. (in Chinese with English abstract)29 相慧,孔祥斌,武兆坤,等. 中国粮食主产区耕地生产能力空间分布特征J. 农业工程学报,2012,28(24):235-244.Xiang Hui, Kong Xiangbin, Wu Zhaokun, et al. Spatial distribution characteristics of potential productivity of arable land in main crop production area in ChinaJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2012, 28(24): 235-244. (in Chinese with English abstract)30 王洪波,程锋,张中帆,等. 中国耕地等别分异特性及其对耕地保护的影响J. 农业工程学报,2011,27(11):1-8.Wang Hongbo , Cheng Feng,Zhang Zhongfan, et al. Differential characteristics of cultivated land grade and its effect on cultivated land protection in ChinaJ. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2011, 27(11): 1-8. (in Chinese with English abstract)31 张青璞. 西部生态脆弱区耕地等别空间分布规律研究D. 北京:中国农业大学,2012.Zhang Qingpu. A Study of the Agricultural Land Grade Distribution Law in West Regions with the Frangible Ecological SystemsD. Beijing:China Agricultural University, 2012. (in Chinese with English abstract)Integration method of accomplishments of arable land from province level to national level based on equal standard crop yieldsKong Xiangbin1,2, Zhang Qingpu1,4, Yun Wenju2,3, Wang Hongbo2,3, Tan Min1,4, Zhao Jing1,4(1. College of Resources and Environmental Sciences, China Agricultural University, Beijing 100193, China; 2. Key Laboratory for Arable Land Quality, Monitoring and Controing of National Ministry of Land Resources, Beijing 100193, China; 3. Land Consolidation and Rehabilitation Center, Ministry of Land and Resources, Beijing 100035, China; 4. Beijing Fang Yuan Da Di Planning and Design Consulting Co.,Ltd., Beijing 100193, China)Abstracts: China, the fastest developing country with the highest population in the world, is undergoing the most rapid industrialization and urbanization in the human history. Thus, the country faces multiple challenges of food insecurity including conversion of prime agricultural land to urban encroachments, limited land resources available for crop production, lack of adequate amounts of fresh water for irrigation and urbanization/industrialization, and increasing risks of climate change and the associated extreme events. Therefore, how to accurately evaluate the potential productivity of arable land at the present technology at national level is very important to identifying yield gap of arable land, and to regulating scientific policy to close yield at national level. China has accomplished arable land grading at national level, and the accomplishments of Chinese national arable land grading at plot level not only pushed the advance in the science and technology of land evaluation, but also provided accurate data basis for management of arable land protection in China based on arable land quantity and quality. The key technique developed in the accomplishment of Chinese national arable land grading was adopted in the arable land grade transferring from provincial to national level. Thus, analyzing the accuracy of the method is very important for identifying the scientific method and providing useful tool for the arable land grading at different levels. The database of arable land grade in 12 provinces and the standard crop yields from a sample of 13,302 at plot level were employed to the compared analysis. The results presented in the article indicated the following:1)the balance transferring method based on the standard crop yields of arable land could fulfill the achievements in arable land grad