Understanding Urban Land Use Changes Using Cellular Automata Model in GIS

1. Analyzing Urban Land Use Changes in Urban EnvironmentsorUsing SLEUTH for Transportation Planning [GIS-T Workshop] Keith C. Clarke Department of Geography UC Santa Barbara

2. Summary • History/Background on SLEUTH • Model theory and operation • Data requirements • Calibration • Outputs • Model use for forecasting • The Role of transportation

3. History/Background of SLEUTH

4. Complex systems theory • Non-linear dynamics • Behavior states: chaos, self-organization, stable • Phase transitions • Emergence • Cellular automata are simple examples of complex systems

5. Cellular Automata • Gridded world • Cells have finite states • Rules define state transitions • Time is incremental • Cells are autonomous, act as agents • Self-replicating machines: Von Neumann • Classic example is Conway’s LIFE

6. Urban Cellular Automata • Cells are pixels • States are land uses • Time is “units”, e.g. years • Rules determine growth and change • Different models have different rule sets • Many models now developed, few tested • Requiem for large scale models (Lee)

7. Background • Urban Dynamics • Regional geographic analysis of temporal spatial data bases • About a dozen applications of three model versions • Retrospective and Analysis • Cellular Automaton Model of Change • Initial set of state conditions (binary to Anderson II) • Change rules • Independent agents • Urban model drives deltatron model

8. Cellular urban modeling • Clarke cellular automaton urban growth model (UGM) • Multiple applications (e.g. San Francisco, Washington/Baltimore) Project Gigalopolis • New applications under way: Chicago, New York, Portland, Philadelphia, MAIA, Mexico City, Santa Barbara • 1998/9 funding made model portable and web-based (USGS: EROS Data Center, EPA Collaboration) • 1999-01 work extends and integrates model with other efforts (LANL and USGS collaboration, NSF Urban Research Initiative, SBECP) • EPA has provided significant input (Ron Matheney and Tom Cathey)

9. Gigalopolis: Project Goal • Use historical data for urban areas to simulate present day urbanization • Simulate using a Cellular Automaton Model (SLEUTH) • Run the model into the future • Simulate alternative futures • Compare across scale and cities • Apply to Urban Dynamics cities

10. Cellular land transition models • Increasingly important methods • Many different models • Link to CA and complex systems theory • Transition probability based • Deltatron changes weighted methods by forcing autocorrelation in change space • Allow modeling, visualization and experiments

11. Model handles land use • So far works at crudest level (Anderson Level 1) • Calibration under way in MAIA and Lower 48 States (GIRAS, MRLC, Loveland) • Needs two LULC layers • Based on the concept of deltatrons • Generates synthetic LU change based on transition matrix and enfored spatial/temporal autocorrelation • Applies CA in change space

12. Land Use Change properties • Driver=urbanization • State probabilities (static, dynamic) • Class magnitudes • Spatial autocorrelation(s)

13. Project Web Site • Set of background materials, e.g. publications • Documentation as web pages in HTML • Source Code for model in C • Version 3.0 now on web for download • Uses utilities and GD GIF libraries • Parallel version requires MPI • Set of sample calibration data demo_city • http://www.ncgia.ucsb.edu/projects/gig/ncgia.html

14. Project Web Site: Shareware C code and Documentation

15. Model Theory and Operation

16. Background • Urban Retrospectives • Geographic analysis • Urban dynamics • Model comparison • Model integration

17. The rules

18. Spontaneous Growth • urban settlements may occur anywhere on a landscape • f (diffusion coefficient, slope resistance)

19. Creation of new spreading centers • Some new urban settlements will become centers • of further growth. • Others will remain isolated. • f (spontaneous growth, breed coefficient, • slope resistance)

20. Organic Growth • The most common type of development • occurs at urban edges and as in-filling • f (spread coefficient, slope resistance)

21. Road Influenced Growth • Urbanization has a tendency to follow lines of transportation • f (breed coefficient, road_gravity coefficient, slope resistance, diffusion coefficient)

22. Behavior Rules T0 T1 spreading center road influenced deltatron spontaneous organic

23. Behavior Rules spreading center road influenced deltatron spontaneous organic T0 T1 f (slope resistance, breed coefficient) f (slope resistance, spread coefficient) f (slope resistance, diffusion coefficient, breed coefficient, road gravity) f (slope resistance, diffusion coefficient) For i time periods (years)

24. Patterns/process of land cover change • Introduction of new land cover type (invasion, diffusion) • Land cover class extension from edges (spread, contagion) • Perpetuation of change (lagged autocorrelation)

25. Deltatron Dynamics: Land cover Delta-space • To/From Transition matrix • Table of land cover class average slopes • Urbanization drives change within the model • Urban (and others) invariant class

26. Deltatrons at work

27. A Deltatron is: • “Bringer of change” (semi-independent agent) • Placeholder of where and what type of land cover transition took place during its lifetime • Tracks how much time has passed since a change has occurred (Lifetime) • Enforces spatial and temporal auto-correlation of land cover transitions by its life cycle

28. Average slope Create delta space For n new urban cells Transition Probability Matrix Deltatron Land Cover ModelPhase 1: Create change Of the two: Find the land class most similar to current slope select random pixel Select two land classes at random spread change change land cover Check the transition probability

29. Deltatron Land Cover ModelPhase 2: Perpetuate change search for change in the neighborhood find associated land cover transitions delta space Transition Probability Matrix • create • deltatrons • impose change in land cover Age or kill deltatrons

30. Data Requirements

31. 1900 1925 1950 1975 2000 • Slope • Land Cover • Excluded • Urban • Transportation • Hillshade

32. Thematic Data Input Consistency Between Data Layers • Hierarchy vs. Definition • More problematic with increase of temporal scope

33. Thematic Data Input Consistency Between Data Layers(cont.) • Consistent data source • Project Specific Documentation

34. ThematicData Input Data resolution • Optimum resolution of data layers is unknown • The SLEUTH can “work” for any data resolution • Tested at 30m to 1km • Roads least realistic at coarse scale

35. Thematic Data Input: Issues • Vertical Integration of Temporal Data Layers • Misregistration produced artificial change • Deurbanization particularly upsetting to model • Road breaks should be avoided

36. UGM Process Flow Data Set Preparation Create Geographic Temporal Database • Source data • historical maps, areal photographs, remotely sensed data, GIS vector/grid data • Select by attribute • urban • transportation • landuse • excluded • slope • Geo-registration • extent (lat, long) • Data type standardization • vector to raster • ArcInfo vector data: LINEGRID or POLYGRID • resolution (rows, columns)

37. Urban Values: 0 = not urban, 0 < n < 255 = urban Roads Values: 0 = not road, 0 < n < 255 = road UGM Process Flow Data Set PreparationImage Format Specifics

38. Landuse: any method can be used Values: Each value matches a given classification value. 1 = urban, 2 = agriculture, 3 = rangeland, etc. Slope: the average percent slope of the terrain is derived from a DEM Values: 0 - 100 UGM Process Flow Data Set PreparationImage Format Specifics

39. Excluded Areas: water bodies and land where urbanization cannot occur. This layer may contain binary data (0 and 99) or ranged values indicating probabilities of exclusion. Values: 0-99. 0 = not excluded, 99 = excluded Background: hillshaded image of region (used only with the graphic version of the model) UGM Process Flow Data Set PreparationImage Format Specifics

40. Final data format must be as a GIF image. ArcInfo: GRIDIMAGE -> TIF xv: TIF -> GIF Create Schedule Files urban.dates roads.dates landuse.dates landuse.classes Naming convention Contents of urban.dates 1930.urban 1950.urban 1970.urban 1990.urban Contents of landuse.classes 0 Unclass UNC 1 Urban URB 2 Agric 3 Range 4 Forest 5 Water EXC 6 Wetland 7 Barren UGM Process Flow Data Set Preparation

41. Thematic Data Input Exclusion Feature Hierarchy and Probability • The exclusion layer • Previously: Binary • static possibility of growth occurring • The Latest: a range (0 - 100) • Enables the exploration of zoning scenarios • e.g.; green zones and urban corridors

42. Calibration

43. Calibration • Most essential element • Ensures realism • Ensures accountability and repeatability • Tests sensitivity • Required for complex systems models • Conducted in Monte Carlo mode

44. The Method • “Brute force calibration” • Phased exploration of parameter space • Start with coarse parameter steps and coarsened spatial data • Step to finer and finer data as calibration proceeds • Good rather than best solution • 5 parameters 0-100 = 101^5 permutations

45. The Problem • Model calibration for a medium sized data set and minimal data layers requires about 1200 CPU hours on a typical workstation • CS calls problem tractability

46. Implementations to date • DEC Alpha • Silicon Graphics (Indy 10000 and O2) • Silicon Graphics Origin 2000 cluster 32 processors: 2GB RAM • Rolla, MS MCMC Beowulf Linux Cluster • Supercomputers (NESC EPA: NC) • Cray C-90 and T3D • Cray T3E-1200

47. past Calibration Predicting the present from the past For n Monte Carlo iterations For n coefficient sets “present”

48. UGM Process Flow UGM Compilation • Download Programs and Data (into a new directory) • contents of downloaded UGM.tar.gz • Clarke Urban Growth Model • Land Cover Deltatron Model • gd libraries • schedule files and calibrate file set to accept demo_city • demo_city data set

49. UGM Process Flow UGM Compilation • Set Up Model and Utilities • gunzip and untar the UGM file • Compile the gd libraries • by entering “make” in the GD subdirectory • In the Model Directory • enter: "make" to compile the model • Type: "grow" • this will begin the program • The user will be prompted for what type of run, output and coeffecient values are desired. • These values are entered into the calibrate file. • test mode • animation vs calibrate • Verify results • compare stats from demo_city with documented results

50. UGM Process Flow UGM Calibration • Phases of calibration • Coarse • Iterations in large increments spanning coefficients’ full range • images 1/4 full size • Fine • Increments are smaller with a more focused coefficient range • images 1/2 full size • Final • The coefficient range should be narrowed to single increments • images are full size