[6] :Kuriakose [9] ( : ) ;Hengl [10] ;Peníěek [11] ; ; ~ ~ m km % -5.6~ ~772.8mm 73

Size: px

Start display at page:

Download "[6] :Kuriakose [9] ( : ) ;Hengl [10] ;Peníěek [11] ; ; ~ ~ m km % -5.6~ ~772.8mm 73"

规钭
7 years ago
Views:

1 GEOGRAPHICAL RESEARCH Vol.31No.10 Oct ( ; ; ) : 533 GIS ( ) ( ) :(1) ( ) ( ) ; ( ) ; (2) ( 1) (3) ( ) : ; ; ; : (2012) [1] [2] [3] [4] [5] [1] [6] [7] [8] : ; : : (2009BAC61B01) : (1981-) yixiangsheng2004@163.com : (1963-) GIS ligs@igsnrr.ac.cn

2 [6] :Kuriakose [9] ( : ) ;Hengl [10] ;Peníěek [11] ; ; ~ ~ m km % -5.6~ ~772.8mm 730~1700mm ; 1 38 ; ( ) [12] [13] ( 5 ) [14] 1 (htp:// (2008)

3 10 : 1795 [15] 407 [16] 100~150cm C ; ; ( m DEM ) Fig.1 SpatialdistributionofsoilprofilesintheThree-RiverHeadwatersRegionofQinghaiProvince SPSS18.0 KolmogorovSmirnov ArcGIS10.0 (GeostatisticsAnalysis) ( ) ( ) [17] (cross-validation) [17] [18] 445

4 (ME) (RMSE) (ASE) [1719] (MSE) (RMSSE) : 0; 0; ; ; 1 1 (ME) (RMSE) ME ;RMSE ;ME RSME ; : ME = n 1 n (V oi -V pi ) RMSE = 1 n i=1 n i=1 槡 2 (V oi -V pi ) (1) V oi V pi n : <10%; 10% < <100%; >100% SPSS 18.0 Kolmogorov-Smirnov (P<0.052-tailed) ( ) ( 1) 1 Tab.1 StatisticalfeaturevaluesofsoildepthintheThree-RiverHeadwatersRegionofQinghaiProvince (cm) (cm) (cm) (cm) ArcGIS10.0 ( 2) 2 X Y Z ;XY ; - - 2

5 10 : Fig.2 Surfacetrendofsoildepth (1) 1 ( ) 2 ( ) ( 2) 2 ( 3a 3b) 2 Tab.2 PredictederrorsofsoildepthbyusingOrdinaryKrigingmethod ME RMSE ASE MSE RMSSE ME ASE MSE RMSE RMSSE 1 ME RMSE 3a 3b

6 Fig.3 Spatialdistributionsofsoildepthbyusingdiferentinterpolationmethods 30.55cm cm 57.54cm 11.64cm; 37.59cm cm 60.96cm 11.24cm

7 10 : 1799 (2) 3 : ( 3) RMSE ASE RMSSE ME MSE ( 3) RMSE ME 3 (C+C 0 ) (C 0 ) (C+C 0 ) (0.791) > (0.727)> (0.653) 0.5 [20] 3 3 Tab.3 Predictederrorsofsoildepthbyusingdiferentmodels ( ) ( ) C0 C+C0 ME RMSE ASE MSE RMSSE (3) ( ) ( ) 4 ME MSE ;RMSSE 1 ;RMSE ASE ME RMSE RMSE ASE RMSSE ME MSE cm cm 39.54cm cm

8 Tab.4 Predictederrorsofsoildepthbyusinggeostatisticsmethods ( ) ( ) C0 C+C0 ME RMSE ASE MSE RMSSE OK SK UK CK OK SK UK CK :OK ;SK ;UK ;CK 34.94cm cm 56.71cm 9.654cm ( 3f); 27.18cm cm 57.43cm 10.25cm ( 3d); 24.90cm cm 57.34cm 12.46cm ( 3c) ( 3c 3f) ME ( 1) ( 1) ( ) RMSE ( ); ME ( 3) RMSE ( 1) 4 ( 1) ( 1) ( 1) ( )( 5) ( 1) 37.49cm 79.31cm 58.40cm 8.76cm ( 3i) ( ) 24.71cm cm 58.94cm 9.66cm; (

9 10 : Tab.5 Predictederrorsofsoildepthbyusingdiferentdeterministicinterpolationmethods ME RMSE ME RMSE ) 12.40cm cm 59.09cm 9.39cm; ( 1) cm cm 58.22cm 28.34cm ( 1) ( 3h) 3g 3j ( 1) ( ) ( 1) ( ) ( 1) ( ) ( 1) ( 6) ( 1) ( ) ME ( ) ( 1) ( ) RMSE ( ) ME ( 1) ( ) RMSE ( ) ( ) ( ) ( 1) ( 3) ( )

10 Tab.6 Predictederrorsofsoildepthbyusingdiferentinterpolationmethods ME RMSE ME RMSE cm 24.90cm 57.34cm 12.45cm ; [21] 4 [22] [23] [24~26] [17] 5~150cm [17]

11 10 : 1803 [2728] [9~1129] α= : (1) (2) ( ) (3) (4) ( 1) (5) ( )

12 : [1] (4):57~59. [2] (4):164~ 167. [3] ( ):8~11. [4] (7):1565 ~1570. [5] LiangXWoodEFLetenmaierDP.SurfacesoilmoistureparameterizationoftheVIC-2L model:evaluation andmodification.globalandplanetarychange (1-4):195~206. [6] (2):161~169. [7] (7):3493~3500. [8] (1):88~92. [9] KuriakoseSLDevkotaSRossiterDetal.Predictionofsoildepthusingenvironmentalvariablesinananthro- pogeniclandscapeacasestudyinthe WesternGhatsofKeralaIndia.Catena (1):27~38. [10] HenglTHeuvelinkGSteinA.Agenericframeworkforspatialpredictionofsoilvariablesbasedonregression- kriging.geoderma (1-2):75~93. [11] PeníekVBorvkaL.Soildepthpredictionsupportedbyprimaryterrainatributes:Acomparisonofmethods. PlantSoilandEnvironment200652:424~430. [12]. : ~327. [13]. : ~262. [14]. ( ). : ~876. [15]. : ~131. [16]. : [17].ArcGIS. : ~428. [18] ZhaoY CShiXZYuDSetal.Soilorganiccarbondensityin HebeiProvinceChina:estimatesanduncertain- ty.pedosphere (3):293~300. [19] (3):132~ 137. [20]. : ~149. [21] SamraJSGilH SAnlaufRetal.GeostatisticalevaluationofsoilsodicityandgrowthofMeliaazedarach Linn.assimultaneousstochasticprocesses.SoilScience (5):361~369. [22] (3):260~265. [23]. GIS. : (2):183 ~188. [24] (5):124~129. [25] (1):47~56. [26] (2):68~72. [27] ZhangX YSuiY YZhangX Detal.SpatialvariabilityofnutrientpropertiesinblacksoilofNortheastChina. Pedosphere (1):19~29. [28] (4):649~658. [29] ZiadatF M.Predictionofsoildepthfromdigitalterraindatabyintegratingstatisticalandvisualapproaches.Pedo- sphere (3):361~367.

13 10 : 1805 Comparisononsoildepthpredictionamong differentspatialinterpolationmethods:acasestudy inthethree-riverheadwatersregionofqinghaiprovince YiXiang-sheng 12 LiGuo-sheng 1 YinYan-yu 34 PengJing-tao 5 (1.InstituteofGeographicSciencesandNaturalResourcesResearchCASBeijing100101China; 2.AgricultureResourceMonitoringStationChineseAcademyofAgricultureEngineeringBeijing China;3.SchoolofGeographicalsciencesSouthwestUniversityChongqing400715China) versedistance WeightedGlobalPolynomialInterpolationLocalPolynomialInterpolation andradialbasisfunction)andgeostatisticsinterpolation methods (Oridnary Kriging Abstract:Basedonthesoildepthdatafrom533soilprofilesintheThree-RiverHeadwaters RegionofQinghaiProvincewiththehelpofGIStechnologythispaperatemptedtopre- dictthespatialdistributionofsoildepthbyusingdeterministicinterpolationmethods(in- SimpleKrigingUniversalKrigingandCo-Kriging).Thenitcomparedthepredictioner- rorsstatisticalcharacteristicsandinterpolationresultsofdiferentinterpolationmethods. Someimportantconclusions wereobtainedfrom thisresearchwhich mainlycontained threeaspectsasfolows.(1)thespatialdistributionofsoildepthusingthefirstorder surfacetrendof Ordinary Kriging wasbeterthanthatusingthesecondordersurface trend.thesphericalmodelintheordinarykriging(firstorder)wasbeterthantheexpo- nentialmodelandgaussianmodel.theordinarykrigingwasthebestofthefourgeosta- tisticsinterpolationmethodsbecauseofitsminimumerrorandaccuratelypredictedresult. (2)TheInverseDistance Weighted(exponent1)methodwasthebestindeterministicin- terpolationmethodsfromthepredictionerrorsandcomprehensivereflectioningeneraland localtrends.(3)incomparisonofthepredictionerrorsandspatialdistributioningeneral andlocaltrendstheordinarykriging (firstorder)ofsphericalmethodwhichhadani- sotropycouldbebesttoreflectthespatialdistributionofsoildepthinthethree-river HeadwatersRegionofQinghaiProvince. Keywords:soildepth;spatialdistribution;spatialinterpolation;Three-RiverHeadwatersRegion

並責成各里幹事下里服勤宣導病媒防治知識, 協助各家戶清除病媒孳生源 ( 積水容器 ), 降低棲群密度, 預防傳染病之發生, 以確保民眾身體健康及居家生活品質訂定每月最後

並責成各里幹事下里服勤宣導病媒防治知識, 協助各家戶清除病媒孳生源 ( 積水容器 ), 降低棲群密度, 預防傳染病之發生, 以確保民眾身體健康及居家生活品質訂定每月最後 541 94.4.6 臺北市文山區都市計畫案通盤檢討主要計畫暨細部計畫案 542 94.5.5 都市計畫道路用地變更為可發展用地免予回饋原則附件三溫泉產業特定專用區都市計畫案召集人本案案情複雜, 且為求審議效益, 委請陳委員武正擔任