应用空间分析技术对浙江省饮用水水质状况进行分析
|
摘要
研究目的
饮用水卫生安全问题一直是政府和人民群众关注的热点,水质的优劣直接影响人民的身体健康。据统计,80%的流行病学资料具有空间属性,同样饮用水数据也有着明显的地理空间分布特征。运用空间分析继续能够利用传统的统计学方法尚未利用的空间信息,为研究者提供一种全新、可靠、科学合理的处理空间信息的方法。本次研究结果,展现了浙江省饮用水水质状况,同时也为政府制定有关饮用水政策提供理论依据。
资料来源与方法
本文的资料来源于2010年浙江省饮用水水质监督监测数据,和国家基础地理信息网数据库中获取的1：400万中国县界电子地图。本次监测的饮用水水样为浙江省范围内的取得有效卫生许可证水厂的出厂水、管网末梢水。出厂水枯水期或丰水期监测一次；管网末梢水为供水区域内按供水人口每2万人设1个采样点,每个采样点每季度监测1次。按《生活饮用水卫生标准》(GB5749-2006)和《生活饮用水标准检验方法》(GB/T5750-2006)进行检验评价。各地饮用水水质监督监测数据汇总、合格率计算和数据变换在EXCEL2003中完成,空间分析在ARCGIS10.0和GS+9.0中完成。
结果
1.饮用水水质情况地区分布图：根据2010年浙江省饮用水水质监督监测数据和浙江省县界电子地图制作成出厂水合格率地区分布图和管网末梢水合格率地区分布图。
2.三维趋势分析：出厂水和管网末梢水在东西和南北方向在均存在趋势,其中管网末梢水在南北方向上浙中高,浙南、浙北低的趋势更显著。
3.变异函数拟合：出厂水块金值C0为0.0095、基台值Co+C为0.2040、块金基台比为0.047、自相关a为0.297、拟合优度r2为0.616,拟合模型较好；管网末梢水块金值C0为0.0799、基台值Co+C为0.1608、块金基台比为0.497、自相关a为2.38、拟合优度r2为0.370,拟合模型一般。
4.kriging插值：出厂水的合格率较高的地区主要在浙西南,较低地区主要在浙东、南沿海地区,插值效果评价指标分别为,估计偏差均数(M-PE)为0.005024、估计偏差标准化均数(MS-PE)为0.01058,估计偏差标化均方根(RMSS-PE)为0.9694,估计偏差均方根(RMS-PE)为0.4477,估计偏差平均标准误(ASE-PE)为0.4624；管网末梢水的合格率较高的地区主要在浙西南和杭州湾附近,较低地区主要在浙南、东沿海地区和浙北地区,插值效果评价指标分别为,估计偏差均数(M-PE)为0.01816、估计偏差标准化均数(MS-PE)为0.04646,估计偏差标化均方根(RMSS-PE)为1.0107,估计偏差均方根(RMS-PE)为0.3187,估计偏差平均标准误(ASE-PE)为0.3152。这说明kriging插值预测是无偏、最优插值。
5.空间自相关分析：经过出厂水和管网末梢水全域Moran'sⅠ和全域G系数分析,只有管网末梢水合格率的Moran'sⅠ系数为0.2865,P<0.05,其余均无统计学意义,说明管网末梢水在整个浙江省区域内存在的正向空间自相关,呈聚集性分布。局域Moran'sⅠ系数和局域Getis系数的Z值检验结果中,出厂水和管网末梢水水质的聚集性表现有非常强的相似性,水质“好”的聚集区在浙西南,遂昌县、龙游县附近区域,水质“差”的聚集区在浙东南沿海,瑞安市、平阳县、苍南县附近区域。
结论
本文应用空间分析技术,直观地显示了浙江省饮用水水质的地理分布,明确了出厂水和管网末梢水水质的聚集性表现有非常强的相似性,水质“好”的聚集区在浙西南,遂昌县、龙游县附近区域,水质“差”的聚集区在浙东南沿海,瑞安市、平阳县、苍南县附近区域,这为政府部门制定相关政策和措施提供了参考信息。
Objective Drinking water safety is always the hotspot the government and people concerned. The water quality will directly influence the people's health. According to statistics,80%of the epidemiological data has spatial attributes, also drinking water data has obvious geographical characteristics of spatial distribution. Spatial analysis can use the spatial information that traditional statistical methods not used yet.It provides researchers with a new, reliable, scientific and reasonable spatial information processing method. The results of this thesis, show the quality of drinking water situation in Zhejiang Province, but also provide a theoretical basis for the government to develop drinking water policy.
Material and methods The drinking water quality monitoring data of Zhejiang Province in2010was collected. And the electronic map was obtained from the electronic map of1:400million Chinese county boundaries in National Fundamental Geographic Information Network database. The finished water and terminal tap water samples of all waterworks with hygiene license in Zhejiang Province formed the monitoring drinking water samples. The finished water was monitored once in dry or flood season. The terminal tap water was monitored as follows:one terminal tap water sampling was set with population of per20,000in water supply area; each sampling was monitored once quarterly."Drinking water health standards (GB5749-2006) and drinking water standard test methods"(GB/T5750-2006) carried out the inspection and evaluation. The drinking water quality monitoring data summary, the qualification rate of computing and data conversion completed in EXCEL2003. Spatial analysis completed in the ARCGIS10.0and GS+9.0.
Result1. The drinking water quality of area maps The finished water and terminal tap water qualification rate area maps produced under the monitoring data of2010drinking water quality in Zhejiang Province and Zhejiang County-border electronic map.
2. Three-dimensional trend analysis The finished water and terminal tap water qualification rate exist trends in the East-West and North-South direction. The trend of higher terminal tap water qualification rate in Central Zhejiang than in other regions in the North-South direction is more pronounced.
3. The fitting results of semivariogram function Finished water:nugget variance is0.0095, sill variance is0.2040,the ratio of nugget variance to sill variance is0.047, the maximum range of spatial autocorrelation is0.297,and goodness of fit is0.616, model fitting is good. Terminal tap water:nugget variance is0.0799, sill variance is0.1608,the ratio of nugget variance to sill variance is0.497, the maximum range of spatial autocorrelation is2.38, and goodness of fit is0.370, model fitting is general.
4. Kriging interpolation Finished water:by distribution map, we find that the finished water qualification rate in the southwest of Zhejiang is higher than in the southeast coastal regions. The evaluation indices of prediction error are the following: Mean is0.005024,Mean Standardized is0.01058, Root Mean Square Standardized is0.9694, Root Mean Square is0.4477, Average Standard Error is0.4624. Terminal tap water:we find that the terminal tap water qualification rate in the southwest of Zhejiang and near the Hangzhou Bay area is higher than in the north and southeast coastal regions. The evaluation indices of prediction error are the following:Mean is0.01816,Mean Standardized is0.04646, Root Mean Square Standardized is1.0107, Root Mean Square is0.3187, Average Standard Error is0.3152.The prediction error of cross validation indicates that the spatial distribution map of the finished water and terminal tap water in Zhejiang is a good fitness.
5. Analysis of spatial autocorrelation After the global Moran's I and the global Getis coefficients analysis of finished water and terminal tap water, only the global Moran's I coefficients of terminal tap water is0.2865and has the statistical significance(P<0.005). Global spatial autocorrelation indicators illustrate that there is cluster with higher qualification rate of terminal tap water in Zhejiang. Through the local Moran's I and the local Getis coefficient analysis, the clustering performance of finished water and terminal tap water is very similar. Regions with higher water qualification rate cluster locate in the southwest of Zhejiang, near Suichang and Longyou County, while regions with lower rate locate in the southeast coast of Zhejiang, near Ruian City, Pingyang and Cangnan County.
Conclusions This thesis shows graphically the geographical distribution of drinking water quality in Zhejiang Province by spatial analysis. It is proved that the clustering performance of finished water and terminal tap water is very similar. Regions with higher water qualification rate cluster locate in the southwest of Zhejiang, near Suichang and Longyou County, while regions with lower rate locate in the southeast coast of Zhejiang, near Ruian City, Pingyang and Cangnan County. So it is beneficial for enacting corresponding policy and taking measures.
饮用水卫生安全问题一直是政府和人民群众关注的热点,水质的优劣直接影响人民的身体健康。据统计,80%的流行病学资料具有空间属性,同样饮用水数据也有着明显的地理空间分布特征。运用空间分析继续能够利用传统的统计学方法尚未利用的空间信息,为研究者提供一种全新、可靠、科学合理的处理空间信息的方法。本次研究结果,展现了浙江省饮用水水质状况,同时也为政府制定有关饮用水政策提供理论依据。
资料来源与方法
本文的资料来源于2010年浙江省饮用水水质监督监测数据,和国家基础地理信息网数据库中获取的1：400万中国县界电子地图。本次监测的饮用水水样为浙江省范围内的取得有效卫生许可证水厂的出厂水、管网末梢水。出厂水枯水期或丰水期监测一次；管网末梢水为供水区域内按供水人口每2万人设1个采样点,每个采样点每季度监测1次。按《生活饮用水卫生标准》(GB5749-2006)和《生活饮用水标准检验方法》(GB/T5750-2006)进行检验评价。各地饮用水水质监督监测数据汇总、合格率计算和数据变换在EXCEL2003中完成,空间分析在ARCGIS10.0和GS+9.0中完成。
结果
1.饮用水水质情况地区分布图：根据2010年浙江省饮用水水质监督监测数据和浙江省县界电子地图制作成出厂水合格率地区分布图和管网末梢水合格率地区分布图。
2.三维趋势分析：出厂水和管网末梢水在东西和南北方向在均存在趋势,其中管网末梢水在南北方向上浙中高,浙南、浙北低的趋势更显著。
3.变异函数拟合：出厂水块金值C0为0.0095、基台值Co+C为0.2040、块金基台比为0.047、自相关a为0.297、拟合优度r2为0.616,拟合模型较好；管网末梢水块金值C0为0.0799、基台值Co+C为0.1608、块金基台比为0.497、自相关a为2.38、拟合优度r2为0.370,拟合模型一般。
4.kriging插值：出厂水的合格率较高的地区主要在浙西南,较低地区主要在浙东、南沿海地区,插值效果评价指标分别为,估计偏差均数(M-PE)为0.005024、估计偏差标准化均数(MS-PE)为0.01058,估计偏差标化均方根(RMSS-PE)为0.9694,估计偏差均方根(RMS-PE)为0.4477,估计偏差平均标准误(ASE-PE)为0.4624；管网末梢水的合格率较高的地区主要在浙西南和杭州湾附近,较低地区主要在浙南、东沿海地区和浙北地区,插值效果评价指标分别为,估计偏差均数(M-PE)为0.01816、估计偏差标准化均数(MS-PE)为0.04646,估计偏差标化均方根(RMSS-PE)为1.0107,估计偏差均方根(RMS-PE)为0.3187,估计偏差平均标准误(ASE-PE)为0.3152。这说明kriging插值预测是无偏、最优插值。
5.空间自相关分析：经过出厂水和管网末梢水全域Moran'sⅠ和全域G系数分析,只有管网末梢水合格率的Moran'sⅠ系数为0.2865,P<0.05,其余均无统计学意义,说明管网末梢水在整个浙江省区域内存在的正向空间自相关,呈聚集性分布。局域Moran'sⅠ系数和局域Getis系数的Z值检验结果中,出厂水和管网末梢水水质的聚集性表现有非常强的相似性,水质“好”的聚集区在浙西南,遂昌县、龙游县附近区域,水质“差”的聚集区在浙东南沿海,瑞安市、平阳县、苍南县附近区域。
结论
本文应用空间分析技术,直观地显示了浙江省饮用水水质的地理分布,明确了出厂水和管网末梢水水质的聚集性表现有非常强的相似性,水质“好”的聚集区在浙西南,遂昌县、龙游县附近区域,水质“差”的聚集区在浙东南沿海,瑞安市、平阳县、苍南县附近区域,这为政府部门制定相关政策和措施提供了参考信息。
Objective Drinking water safety is always the hotspot the government and people concerned. The water quality will directly influence the people's health. According to statistics,80%of the epidemiological data has spatial attributes, also drinking water data has obvious geographical characteristics of spatial distribution. Spatial analysis can use the spatial information that traditional statistical methods not used yet.It provides researchers with a new, reliable, scientific and reasonable spatial information processing method. The results of this thesis, show the quality of drinking water situation in Zhejiang Province, but also provide a theoretical basis for the government to develop drinking water policy.
Material and methods The drinking water quality monitoring data of Zhejiang Province in2010was collected. And the electronic map was obtained from the electronic map of1:400million Chinese county boundaries in National Fundamental Geographic Information Network database. The finished water and terminal tap water samples of all waterworks with hygiene license in Zhejiang Province formed the monitoring drinking water samples. The finished water was monitored once in dry or flood season. The terminal tap water was monitored as follows:one terminal tap water sampling was set with population of per20,000in water supply area; each sampling was monitored once quarterly."Drinking water health standards (GB5749-2006) and drinking water standard test methods"(GB/T5750-2006) carried out the inspection and evaluation. The drinking water quality monitoring data summary, the qualification rate of computing and data conversion completed in EXCEL2003. Spatial analysis completed in the ARCGIS10.0and GS+9.0.
Result1. The drinking water quality of area maps The finished water and terminal tap water qualification rate area maps produced under the monitoring data of2010drinking water quality in Zhejiang Province and Zhejiang County-border electronic map.
2. Three-dimensional trend analysis The finished water and terminal tap water qualification rate exist trends in the East-West and North-South direction. The trend of higher terminal tap water qualification rate in Central Zhejiang than in other regions in the North-South direction is more pronounced.
3. The fitting results of semivariogram function Finished water:nugget variance is0.0095, sill variance is0.2040,the ratio of nugget variance to sill variance is0.047, the maximum range of spatial autocorrelation is0.297,and goodness of fit is0.616, model fitting is good. Terminal tap water:nugget variance is0.0799, sill variance is0.1608,the ratio of nugget variance to sill variance is0.497, the maximum range of spatial autocorrelation is2.38, and goodness of fit is0.370, model fitting is general.
4. Kriging interpolation Finished water:by distribution map, we find that the finished water qualification rate in the southwest of Zhejiang is higher than in the southeast coastal regions. The evaluation indices of prediction error are the following: Mean is0.005024,Mean Standardized is0.01058, Root Mean Square Standardized is0.9694, Root Mean Square is0.4477, Average Standard Error is0.4624. Terminal tap water:we find that the terminal tap water qualification rate in the southwest of Zhejiang and near the Hangzhou Bay area is higher than in the north and southeast coastal regions. The evaluation indices of prediction error are the following:Mean is0.01816,Mean Standardized is0.04646, Root Mean Square Standardized is1.0107, Root Mean Square is0.3187, Average Standard Error is0.3152.The prediction error of cross validation indicates that the spatial distribution map of the finished water and terminal tap water in Zhejiang is a good fitness.
5. Analysis of spatial autocorrelation After the global Moran's I and the global Getis coefficients analysis of finished water and terminal tap water, only the global Moran's I coefficients of terminal tap water is0.2865and has the statistical significance(P<0.005). Global spatial autocorrelation indicators illustrate that there is cluster with higher qualification rate of terminal tap water in Zhejiang. Through the local Moran's I and the local Getis coefficient analysis, the clustering performance of finished water and terminal tap water is very similar. Regions with higher water qualification rate cluster locate in the southwest of Zhejiang, near Suichang and Longyou County, while regions with lower rate locate in the southeast coast of Zhejiang, near Ruian City, Pingyang and Cangnan County.
Conclusions This thesis shows graphically the geographical distribution of drinking water quality in Zhejiang Province by spatial analysis. It is proved that the clustering performance of finished water and terminal tap water is very similar. Regions with higher water qualification rate cluster locate in the southwest of Zhejiang, near Suichang and Longyou County, while regions with lower rate locate in the southeast coast of Zhejiang, near Ruian City, Pingyang and Cangnan County. So it is beneficial for enacting corresponding policy and taking measures.
引文
[1]刘昌明,何希吾.我国21世纪上半叶水资源需求分析[J].中国水利,2000(1)；19-20.
[2]陈志恺.21世纪中国水资源持续开发利用问题[J].中国工程科学,2000,2(3)：7-11.
[3]任金法.我国城乡饮水现状及重大饮用水污染状况分析[J].中国初级保健,2009,23(6)：7-11.
[4]王强,赵月朝等.1996-2006年我国饮用水污染突发公共卫生事件分析[J].环境与健康杂志,2010,27(4).
[5]郭向楠,张勇.2007-2008年中国城乡饮用水源突发污染事件统计及分析[J].安全与环境学报,2009,9(3)：183-192
[6]张岚,王丽等.国际饮用水水质标准现状及发展趋势[J]环境与健康杂志,2007,24(6)：451-453.
[7]GB5749-2006,生活饮用水卫生标准[s].
[8]张剑铭,蔡建民.浙江省1997～1999年饮用水卫生检测结果[J].浙江预防医学,2001,13(11)：1-2.
[9]曾瑞胜,杨新新,姜伟.浙江省农村饮用水水质状况分析[J].中国农村水利水电,2004(4)：19-21
[10]武先锋,陶勇.GIS在饮水与健康领域中的应用及开发[J].国外医学(卫生学分册),2005,32(5)：285-289.
[11]Carlos CS.Use of geographic information system in epidemiology(GIS—Epi).
[12]GB/T 5750-2006,生活饮用水标准检验方法[s].
[13]Monica Partt在ArcGIS中处理Excel表格[J].冯磊,译ArcGIS World,2010(5):45-50.
[14]吴秀芹,张洪岩,李瑞改,等.ArcGIS9地理信息系统应用与实践(下册)[M].北京：清华大学出版社,2007.
[15]汤国安,杨昕ArcGIS地理信息系统空间分析实验教程[M].北京：科学出版社,2010.
[16]Kitron U,Kazmierczak J.Spatial analysis of the distribution of Lyme disease in Wisconsin[J]. American Journal of Epidemiology,1997,145(6):558-566.
[17]周国法,徐汝梅.生物地理统计学一生物种群时空分析的方法及其应用[M].北京：科学出版社,1998.
[18]王劲峰.空间分析[M].北京：科学出版社,2006.
[19]Lai D.Geostatistical analysis of Chinese cancer morbidity:Variogram,Kriging and Beyond[J].Journal of Data Science,2004,2(2):177-193.
[20]薛付忠,王洁贞,王发银,等.疾病空间分布的变异函数模型及其应用[J].山东医科大学学报,2001,39(1)：27-31.
[21]薛付忠,王洁贞,王振光,等.疾病空间异质性定量分析方法及应用[J].山东大学学报(医学版),2002,40(6)：485-488.
[22]Fabrice Carrat,Alain Jacques Valleron.Epidemiologic Mapping Using the "Kriging" Method:Application to an Influenza-like Illness Epidemic in France[J].Am J Epidemiol,1992,135(11):1293-1300.
[23]Stytz MR,Parrott RW.Using Kriging for 3D medical imaging[J].Comput Med Imaging Graph,1993,17(6):421-442.
[24]王政权.地统计学及在生态学中的应用[M].北京：科学出版社,1999.
[25]Anselin L.Local indicators of spatial association-LISA[J].Geographical Analysis,1995,27(2):93-115.
[26]Kleinsehmidt L,Bagayoko M,Clarke GPY,et al.A spatial statistical approach to malaria mapping[J].Intemational Journal of Epidemiology,2000,29(2):355-361.
[27]浙江省统计局.2010年浙江省统计年鉴http://www.zj.stats.gov.cn/zjtj2010/indexch.htm,2011.9.30
[28]浙江省环境保护厅.2010年浙江省环境状况公报http://www.zjepb.gov.cn/root14/auto505/201106/t20110603_96078.html, 2011.9.30
[29]胡向敏.对温州市农民饮用水工作的思考[J].浙江水利科技,2009(5)：57-58.
[30]周鑫根.浙江省城乡一体化供水体系规划研究[J].给水排水,2006,32(1).
[1]Carlos CS.Use of geographic information system in epidemiology(GIS——Epi).
[2]武先锋,陶勇.GIS在饮水与健康领域中的应用及开发[J].国外医学卫生学分册,2005,32(5)：285—289.
[3]Haining RP,Wise SM, Ma J. Exploratory spatial data analysis in a geographic information system environment[J]. The Statistician,1998,47:457-469.
[4]Berke O.Exploratory spatial relative risk mapping[J].Preventive Veterinary Medicine,2005,71(3):173-182.
[5]WU Jilei,WANG Jinfeng,MENG Bin, et al. Exploratory spatial data analysis for the identification of risk factors to birth defects[J].BMC PublicHealth,2004,4:23.
[6]Thomson MC,Connor SJ.Predicting malaria infection in Gambian children from satellite data and bed net use survey:the importance of spatial correlation in the interpretation of results[J].Am J Trop Med Hyg,1999,61:2-8.
[7]Fabrice Carrat,Alain Jacques Valleron.Epidemiologic Mapping Using the "Kriging" Method:Application to an Influenza-like Illness Epidemic in France[J].Am J Epidemiol,1992,135(11):1293-1300.
[8]Stytz MR,Parrott RW.Using Kriging for 3D medical imaging[J].Comput Med Imaging Graph,1993,17(6):421-442.
[2]陈志恺.21世纪中国水资源持续开发利用问题[J].中国工程科学,2000,2(3)：7-11.
[3]任金法.我国城乡饮水现状及重大饮用水污染状况分析[J].中国初级保健,2009,23(6)：7-11.
[4]王强,赵月朝等.1996-2006年我国饮用水污染突发公共卫生事件分析[J].环境与健康杂志,2010,27(4).
[5]郭向楠,张勇.2007-2008年中国城乡饮用水源突发污染事件统计及分析[J].安全与环境学报,2009,9(3)：183-192
[6]张岚,王丽等.国际饮用水水质标准现状及发展趋势[J]环境与健康杂志,2007,24(6)：451-453.
[7]GB5749-2006,生活饮用水卫生标准[s].
[8]张剑铭,蔡建民.浙江省1997～1999年饮用水卫生检测结果[J].浙江预防医学,2001,13(11)：1-2.
[9]曾瑞胜,杨新新,姜伟.浙江省农村饮用水水质状况分析[J].中国农村水利水电,2004(4)：19-21
[10]武先锋,陶勇.GIS在饮水与健康领域中的应用及开发[J].国外医学(卫生学分册),2005,32(5)：285-289.
[11]Carlos CS.Use of geographic information system in epidemiology(GIS—Epi).
[12]GB/T 5750-2006,生活饮用水标准检验方法[s].
[13]Monica Partt在ArcGIS中处理Excel表格[J].冯磊,译ArcGIS World,2010(5):45-50.
[14]吴秀芹,张洪岩,李瑞改,等.ArcGIS9地理信息系统应用与实践(下册)[M].北京：清华大学出版社,2007.
[15]汤国安,杨昕ArcGIS地理信息系统空间分析实验教程[M].北京：科学出版社,2010.
[16]Kitron U,Kazmierczak J.Spatial analysis of the distribution of Lyme disease in Wisconsin[J]. American Journal of Epidemiology,1997,145(6):558-566.
[17]周国法,徐汝梅.生物地理统计学一生物种群时空分析的方法及其应用[M].北京：科学出版社,1998.
[18]王劲峰.空间分析[M].北京：科学出版社,2006.
[19]Lai D.Geostatistical analysis of Chinese cancer morbidity:Variogram,Kriging and Beyond[J].Journal of Data Science,2004,2(2):177-193.
[20]薛付忠,王洁贞,王发银,等.疾病空间分布的变异函数模型及其应用[J].山东医科大学学报,2001,39(1)：27-31.
[21]薛付忠,王洁贞,王振光,等.疾病空间异质性定量分析方法及应用[J].山东大学学报(医学版),2002,40(6)：485-488.
[22]Fabrice Carrat,Alain Jacques Valleron.Epidemiologic Mapping Using the "Kriging" Method:Application to an Influenza-like Illness Epidemic in France[J].Am J Epidemiol,1992,135(11):1293-1300.
[23]Stytz MR,Parrott RW.Using Kriging for 3D medical imaging[J].Comput Med Imaging Graph,1993,17(6):421-442.
[24]王政权.地统计学及在生态学中的应用[M].北京：科学出版社,1999.
[25]Anselin L.Local indicators of spatial association-LISA[J].Geographical Analysis,1995,27(2):93-115.
[26]Kleinsehmidt L,Bagayoko M,Clarke GPY,et al.A spatial statistical approach to malaria mapping[J].Intemational Journal of Epidemiology,2000,29(2):355-361.
[27]浙江省统计局.2010年浙江省统计年鉴http://www.zj.stats.gov.cn/zjtj2010/indexch.htm,2011.9.30
[28]浙江省环境保护厅.2010年浙江省环境状况公报http://www.zjepb.gov.cn/root14/auto505/201106/t20110603_96078.html, 2011.9.30
[29]胡向敏.对温州市农民饮用水工作的思考[J].浙江水利科技,2009(5)：57-58.
[30]周鑫根.浙江省城乡一体化供水体系规划研究[J].给水排水,2006,32(1).
[1]Carlos CS.Use of geographic information system in epidemiology(GIS——Epi).
[2]武先锋,陶勇.GIS在饮水与健康领域中的应用及开发[J].国外医学卫生学分册,2005,32(5)：285—289.
[3]Haining RP,Wise SM, Ma J. Exploratory spatial data analysis in a geographic information system environment[J]. The Statistician,1998,47:457-469.
[4]Berke O.Exploratory spatial relative risk mapping[J].Preventive Veterinary Medicine,2005,71(3):173-182.
[5]WU Jilei,WANG Jinfeng,MENG Bin, et al. Exploratory spatial data analysis for the identification of risk factors to birth defects[J].BMC PublicHealth,2004,4:23.
[6]Thomson MC,Connor SJ.Predicting malaria infection in Gambian children from satellite data and bed net use survey:the importance of spatial correlation in the interpretation of results[J].Am J Trop Med Hyg,1999,61:2-8.
[7]Fabrice Carrat,Alain Jacques Valleron.Epidemiologic Mapping Using the "Kriging" Method:Application to an Influenza-like Illness Epidemic in France[J].Am J Epidemiol,1992,135(11):1293-1300.
[8]Stytz MR,Parrott RW.Using Kriging for 3D medical imaging[J].Comput Med Imaging Graph,1993,17(6):421-442.