人文gis課程課件_第1頁
人文gis課程課件_第2頁
人文gis課程課件_第3頁
人文gis課程課件_第4頁
人文gis課程課件_第5頁
免費預覽已結(jié)束,剩余27頁可下載查看

下載本文檔

版權說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權,請進行舉報或認領

文檔簡介

1、Function Fittings by Regressions and Application in Analyzing Urban and Regional Density Patterns(Chapter 6, Project 6)OutlineTheories for Urban & Regional Density PatternsEmpirical Models for Monocentric ModelsStudy on BeijingFunction Fittings for Polycentric ModelsStudy on ShenyangCase Study 6: An

2、alyzing Urban Density Patterns in Chicago RegionTheory I: Urban Economic ModelClarks (1951) observation of urban density patternAlonsos (1967) urban land use theory The monocentric model by Mills (1972) & Muth (1969)Max U = U(h,x), s.t. y = phh+pxx+tr(1/ph) ph/r = -t ph=p0e-trFrom land price to popu

3、lation densityBasic concept tradeoff of commute cost vs. living spaceLand Price (ph)rCBDLand Consumption (h)rCBDrCBDPopulation Density (D)Theory II: Gravity-Based Models for Urban DensityWang and Guldmann (1996)Population at a location is proportional to its potential (location advantage)xi = j (xjd

4、ij-b) for all i=1,2, Higher population density near CBD is due to better accessibility shaped by transportation network (e.g., Chicago)Also see the Garin-Lowry Model in Wang (1998)Road Network(Highway) inChicagoRegion Theories for Regional StructureVon Thnens (1826) rural land use theorySpatial equi

5、librium model by Wang and Guldmann (1997)City produces industrial goodsRural area produces agricultural goodsBoth are sold in the cityWage, population density (labor input) & land price vary in distance from the cityUtility is equalized everywhereWage rateDevelopedDevelopingLand rentAgricultural yie

6、ldPopulation densityEmpirical Models (Monocentric)Linear: D = a-brNegative exponential: D = D0e-brlnD = a-brlogarithmic : D = a-b(lnr) Power (log-linear): D = D0r-blnD = a-b(lnr)All are bivariateExtensions to Monocentric:CBD CraterSuburban Bumps2. Newlings model:lnD = a+br-cr23. Cubic Spline Functio

7、n:D = a+b(x-x0)+c(x-x0)2 +d0(x-x0)3+ di(x-xi)3Zi*1. Tanner-Sherratt model: lnD = a-br2 Linear Regressions (Monocentric)ModelsFunction used in regressionOriginal functionX var(s)Y varRestrictionsLinearDr = a+brSamerDrnoneLogarithmic Dr = a+blnrSamelnrDrr0Power lnDr = A+blnr Dr = arblnrlnDrr0 & Dr0Exp

8、onential lnDr = A+brDr = aebrrlnDrDr0Tanner-SherrattlnDr = A+br2r2lnDrDr0NewlingslnDr = A+b1r+b2r2r , r2lnDrDr0Two Statistical IssuesNon-randomness of sampling (more high-density small tracts)Weighted regression Non-linear regression vs. linearized log-transformationD = D0e-br weights all equal abso

9、lute errors equally lnD = a-br weights equal percentage errors equally Both use non-linear regressions in SAS or SPSSBeijing 1982 Density vs. DistanceRegional growth patternsBeijing 1990 Density vs. DistanceRegressions in Beijing (non-weighted)FunctionForms19821990abR2abR2D=a-br37238-22450.48434444-

10、19440.523D=D0e-br70949-0.2290.67153938-0.1680.649D=a-blnr55448-195900.63949529-164150.634D=D0r-b85385-0.8880.63467071-0.7280.593Regressions in Beijing (weighted)Function Forms19821990ababD=a-br22373-121522984-1217D=D0e-br69907-0.26256772-0.212D=a-blnr42883-1501642011-14368D=D0r-b98108-1.17782066-1.0

11、351982-90 Population Change %General Trend: Suburbanization19821990Density (D)Distance (r)Polycentric Models(a) Substitutable influences from centers only the nearest center matters(b) Complementary influences from centers access to all centers is necessary(c) Between (a) & (b)(d) Access to CBD and

12、the nearest subcenter is valued (central-place theory)(a)(b)(d)(c)CBDCBDCBDCBDEmpirical Models (Polycentric)(a) D = D0e-br within each subarea(b) lnD = a+b1r1+b2r2+(c) D = a1exp( b1r1)+ a2exp( b2r2)+(d) lnD = a+ b1rCBD+b2rsubcPolycentric Assumptions and Corresponding Functions LabelAssumptionModel (

13、exponential as an example)X variablesSample Estimation method(A)Only access to the nearest center is neededDistances ri from the nearest center i (1 variable)Areas in a subregion iLinear regression1(B)Access to all centers is necessary (multiplicative effects)Distances from each center (n variables

14、ri)All areasLinear regression(C)Access to all centers is necessary (additive effects)Distances from each center (n variables ri)All areasNonlinear regression(D)Access to CBD and the nearest center is neededDistances from the major and nearest center (2 variables)All areasLinear regression2Polycentri

15、c Structure in ShenyangWang and Meng (1999), based on 1982 & 1990 Population Census DataPoor fitting power of monocentric modelsR2=0.258, 0.140 for negative exponential in 1982 and 1990 respectively Shenyang 1982 Density ContourShenyang 1990 Density ContourRegression (Shenyang): D = aiexp( biri)Mult

16、i-Centers1982 (R2=0.721)1990 (R2=0.678)aibiaibiA27232*(3.54)-0.8749*(-1.96)13553(1.68)-0.1287(-1.75)B54894*(10.11)-0.4236*(-5.54)51421*(9.66)-0.4069*(-4.85)C35944*(4.54)-0.7593*(-2.92)23374*(3.47)-0.5774(-1.70)D25412*(3.11)-0.7156*(-1.97)30245*(3.31)-1.1622*(-2.15)E42758*(6.95)-0.4225*(-3.74)44006*(

17、6.00)-0.6082*(-3.64)Case Study 6 (Part 1): Monocentric Density Functions in Chicago (Tracts)extracting study area and CBD location population density surface modelingcomputing distances between tract centroids and CBD in ArcGISusing the regression tool in Excel to run simple linear regressionsusing

18、the trendline tool in Excel for generating graphs and function fittings nonlinear and weighted regressions in SAS Case Study 6 (Part 2): Polycentric Density Functions in Chicago (Tracts)Computing distances between tract centroids and their nearest centers in ArcGIS test the polycentric assumptions (A) and (D) Computing distances between tract centroids and each center in ArcGIS test the polycentric assumptions (B) and (C). Fitting polycentric functions in SAS Defining job centers by GIS surface modelingSpatial approach to commuting: Job-housing balance ratio ap

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預覽,若沒有圖紙預覽就沒有圖紙。
  • 4. 未經(jīng)權益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負責。
  • 6. 下載文件中如有侵權或不適當內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準確性、安全性和完整性, 同時也不承擔用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

評論

0/150

提交評論