Working with Elevation Data Using Mosaic Datasets & Image Services. Peter Becker

Transcription

1 Working with Elevation Data Using Mosaic Datasets & Image Services Peter Becker

2 OUTLINE Uses or Elevation data Requirements ArcGIS 10 capabilities - Mosaic Datasets - ArcGIS Server Best Practices Worklow or Elevation Data - Data sources, structures, services, NoData, more Web applications using Image Services LiDAR & Terrain support

3 Uses o Elevation As Elevation 3D Visualization Shaded Relie cartographic Aspect Agriculture Slope Land subsidence, o-road mobility analysis Proiles Planning pipelines, drainage Viewshed Visibility analysis Orthorectiication Dierence/Volume Tree Height Change Ground movement

4 Traditional method o Managing Elevation Data Management - Project wise - Separate datasets Analysis - Project wise - Merge together required sources

5 World Elevation Elevation or the complete globe Multi Source - GTOPO, SRTM, - USGS NED (1 and 1/3 arcsecond) - Lidar or sample areas - EGM2008 Geoid model Services - Elevation Orthometric & Ellipsoidal - Hill Shade, Slope, Aspect, Shaded Relieve Tasks - Proile, Viewshed, Contour

6 Demo: World Elevation Service

7 User Requirements Visualization - Hillshade, Slope, Aspect Directly using Elevation - Othophoto generation, Contours, Viewshed - Application needs elevation (or slope, ) Analysis Results - Visibility, Viewshed, Contours - User only gets results Data download & export These requirements can best be achieved using Image Services

8 Data Management Requirements Single Service rom multiple sources - Simpliies data management and dissemination Create Derived Service or dierent - Visual representations - Orthometric vs Ellipsoidal - Surace Elev. (DSM) vs. Ground Elev, (DEM) Set up server side analysis services

9 ArcGIS 10 Provides A Complete Geospatial Platorm Integrating Geospatial Knowledge From Many Sources Mobile Web Discover Create Manage Visualize Analyze Collaborate Cloud Enterprise Local Desktop Making GIS Available to Everyone

10 Manage, Disseminate, Visualize, Analyze Imagery is Core to GIS ArcGIS enables you to: Manage, Disseminate, Visualize and Analyze all orms o imagery Platorm or complete Imagery Solutions Mosaic Datasets are the optimum model o managing and serving imagery and rasters Maximizing the Value o Imagery

11 Managing & Serving Elevation Data

12 ArcGIS For Image Data Management Storage, Catalog, Metadata & Process Workstation User What do I have? How can I easily work with it? Organizations with collections o processed imagery How can I server my elevation data to multiple users? Enterprises collecting new imagery How do I process and serve new elevation that we acquire? Catalog all available data Make it quickly accessible in the required orm

13 Mosaic Dataset Optimum Model or Image Data Management Within ArcGIS Desktop (Editor/Ino) Quickly Catalog - All raster datasets including elevation - Imagery rom dierent sensors Deine In Geodatabase - Metadata - Processing to be applied - Deault viewing rules Access In all ArcGIS applications - As Image - Dynamic Mosaic, Processed on-the-ly - As Catalog - Footprints, Detailed metadata

14 Dynamic Mosaicking Merge multiple sources Projections Formats Bit Depths Pixel Sizes User deined ordering On-The-Fly Processing Process image as accessed Stretch, Extract Bands Clip, Mask Reproject, Orthorectiy, Pan Sharpen Vegetation Index, Classiy Shaded Relie, Slope, Aspect Color Correction t1 t2 t3 t4 t5

15 Data Sources GTOPO 30-arc second (1km) SRTM Shuttle Mission Topography Mission - 3-arc sec (90m) ASTER GDEM 1-arc sec (30m) NED 1-arc sec (30m) & 1/3-arc sec (10m) Intermap, SPOT, other. Photogrammetry Correlation or point capture LiDAR Bathymetry Sonar From Contour, vector sources Varying Horizontal and Vertical accuracies (LE90 CE90)

16 Metadata Should be obtained or all sources Horizontal projection and datum Vertical datum and unit Horizontal accuracy as CE90, Vertical as LE90 Ground or Surace? Data source NoData deinition Ground sample distance Data raw or resampled?

17 Data Structures Gridded Rasters - GRID, TIF, FLT, ASCII DEM, IMG, BAG*, HRE -. - INTEGER vs. FLOAT Irregular - Terrain, LAS, MG4 LiDAR, BAG, ASCII XYZ, 3D SHP -

18 Raster / Grid For Elevation; Generally a Derivative Advantages - Simplest representation - Highly scalable - Fast access Disadvantages - Loss o original point data - Require multiple rasters or dierent classiications, attributes

19 Pre-Processing Elevation data in raster ormat Do NOT re-project! (will be done OTF i required) Optimum ormat TIF w/ LZW compression Some ormats (ASCII DEM, GRID) should be converted Special ormats: HDF, NETCDF. May be better to convert Larger datasets (NCols>5000), better to have pyramids (OVR) I NCols > 5000 and not tiled, consider reormatting JPEG2000 possible, but w/ decompression cost penalty Terrain/LiDAR processing discussed later

20 Mosaic Datasets and Services to Create Orthometric Ground Elevation

21 Mosaic Datasets and Services to Create Orthometric Ground Elevation Hillshade Relieshade Hillshade Slope Aspect

22 Mosaic Datasets and Services to Create Orthometric Ground Elevation Surace Elevation - Hillshade Relieshade Hillshade Slope Aspect Surace Height

23 Mosaic Datasets and Services to Create Orthometric Ground Elevation Surace Elevation - Hillshade Relieshade Hillshade Slope Aspect Surace Height Ground MSL Orthometric Height

24 Mosaic Datasets and Services to Create Orthometric Ground Elevation Surace Elevation - Hillshade Relieshade Hillshade Slope Aspect Surace Height Ground MSL Ellipsoid Ellipsoidal Height Orthometric Height

25 Mosaic Datasets and Services to Create Orthometric Ground Elevation Surace Elevation - Hillshade Relieshade Hillshade Slope Aspect Surace Height Ground MSL Ellipsoid Ellipsoidal Height Orthometric Height Geoid Undulation

26 Mosaic Datasets and Services to Create Orthometric Ground Elevation Surace Elevation - Hillshade Relieshade Hillshade Slope Aspect Geoid Undulation Eg EGM Surace Height Ellipsoidal Ground Elev. Ground MSL Ellipsoid Ellipsoidal Height Orthometric Height Geoid Undulation

27 Mosaic Datasets and Services to Create Orthometric Ground Elevation Surace Elevation - Hillshade Relieshade Hillshade Slope Aspect Geoid Undulation Eg EGM Processing Viewshed Contour Proile Surace Height Ellipsoidal Ground Elev.

28 Mosaic Dataset Design Create Master Mosaic Dataset or Orthometric Ground Elevation - Projection (or management and overviews) World Mercator? - Ensure type = loat Create suitable Metadata attributes - Horiz_CE90, Vert_LE90, Source, Best Separate Mosaic Dataset or - Surace (e.g. LiDAR First Return) Create Reerence / Derived Mosaic Datasets

29 Data Ingest Use suitable Raster Type <Raster Dataset> Consider create new Mosaic Dataset or each source, QC and then add to master Convert units when necessary (scale/oset)

30 Demo Create Mosaic Datasets and Ingest Data

31 NoData Pixels / Areas with No Value Use NoData Value Raster Property Use NoData Mask Some image ormats Deine Mask Value or Range - Functions Deine Footprint - Build Footprint tool - Recommended Options or Oceans/Seas - NoData - 0 (Global dummy image with value=0) - Bathymetric

32 Demo NoData

33 Mosaic Dataset Properties Set Best = most accurate (Vert_LE90 or LoPS) on top LZ77 compression or transmission Set allowable mosaic methods to: ByAttribute, Locked Set MinPS = 0 or all datasets

34 Overviews Deine provide ast access to small scales Can be generated rom primary data Advantageous to use global data sources such as SRTM and GTOPO to mitigate the requirement or creating overviews Review NoData values in overviews

35 Demo Mosaic Dataset Properties

36 Creating Reerence Mosaic Datasets Hill Shade, Slope, Aspect - Add Function - Change Compression or transmission (JPEG) - * there exists also a Derived method Geoid - Need Geoid Undulation model eg EGM As single ile or Mosaic Dataset - Ellipsoid height = Optometric + Geoid Undulation Height - Height = (First Return Bare Earth) - =

37 Demo Reerence Mosaic Datasets

38 Optimization & Maintenance New data can be added as required to master Reerence Mosaic Datasets are automatically updated Consider optimizing ormats I a lot o Nodata Areas - Consider to generate tiles to minimize NoData processing

39 Applications using Elevation based Image Services Serve Visualization - Direct use o Hillshade, Slope, Aspect Directly Using Elevation - Client downloads source data or local processing - Recommendation: Try to minimize this usage mode Serve Analysis Results - Tools to perorm on-demand analysis on server - Viewshed, Proile, Contours In any o 3 uses above, client application can be: - ArcGIS (or urther technical analysis) - Other web apps (ArcExplorer, custom apps, etc.)

40 GeoProcessing with Image Services Requirement: Perorm analysis on multi-resolution data Example: Calculate viewshed, contours Options: - User download ull resolution source elevation data - Need client capable o processing - Too expensive in data transer - Publish GeoProcessing service - Accessible as a service - Requires server to extract required pixels rom image Service - What Cell Size? - Impractical to do Viewshed o Himalayas using 2t spacing

41 GeoProcessing Image Service Options MakeImageServiceLayer - Initialization will look to complete service - Set MaxRows/Cols = Extent/Base PixelSize - Input to tool is URL - Sampling = Nearest Neighbor Resample - Server read & resample only what is required or Analysis - Input to tool is image service layer in ArcMAP document - Sampling method can be deined Download - Can avoid sampling - Query source and download required data (with clipping) - Merge sources according to user s rules - More complex to process, download may be slow

42 Viewshed region o interest Viewshed? x

43 Viewshed region o interest

44 Viewshed region o interest 1000 pixels 1000 pixels

45 Viewshed region o interest x

46 Demo GeoProcessing with Image Services

47 Terrain Dataset Multi-resolution surace created rom measurements stored in eature classesc MassPoints (LiDAR), Breaklines, Ppot heights, Polygons, Stored in the geodatabase Schema: Deines eature class participation On-The-Fly TIN Multi-resolution Highly scalable Attributes Points and Breaklines Editable, Versioned Terrain Pyramids High Resolution Medium Resolution Low Resolution

48 Rasterizing Terrain Easily convert Terrain datasets to Raster - Size based on Point spacing - TIF with LZW - NoData / Boundaries Considerations - Have overlap in tiles (20 pixels)

49 LiDAR (Light Detection and Ranging) Laser based High Density, High Accuracy Partially penetrates canopy LAS ASPRS Include classiication

50 Rasterizing LiDAR Import LAS to Multi-point; Then, two methods: (1) Raster via Terrain - Point File Inormation (Avg. point spacing) - Terrain Wizard; 2 Terrains or First and Last (Ground) Return - NoData holes illed by TIN - QC and Edit Terrain to Raster (previous slide) (2) Interpolate Multi-point directly to Raster - Point to Raster; Set Point Set cell size to 4x avg. point spacing - Filter to remove holes: Con(IsNull("INPUTRASTER"), FocalStatistics("INPUTRASTER", NbrRectangle(3,3, "CELL"), "MEAN", "DATA"), "INPUTRASTER") 3 rd Party Tools - Eg LP360 (QCoherent) Considerations - Have overlap in tiles (20 pixels)

51 Including Terrain and Lidar Rasters Mosaic Dataset Image Service View and Download View & Export

52 Including Terrain and Lidar Rasters Mosaic Dataset Image Service View and Download View & Export Spatial Analysis Desktop - Volume Estimates Viewshed Contour Proile

53 Including Terrain and Lidar Rasters Mosaic Dataset Image Service GeoProcessing Service View and Download View & Export Spatial Analysis Desktop - Viewshed Contour Proile Viewshed Contour Proile Volume Estimates

54 Including Terrain and Lidar Rasters Mosaic Dataset Image Service LAS iles Rasters GeoProcessing Service View and Download View & Export Spatial Analysis Desktop - Viewshed Contour Proile Viewshed Contour Proile Volume Estimates

55 Including Terrain and Lidar Rasters Mosaic Dataset Image Service Rasters LAS iles View and Download Constraints Terrain Dataset GeoProcessing Service View & Export Terrain Analysis Spatial Analysis Desktop Viewshed Contour Proile - Edit Hydrology Viewshed Contour Proile Volume Estimates

56 Including Terrain and Lidar Rasters Mosaic Dataset Image Service Rasters LAS iles View and Download Constraints Terrain Dataset GeoProcessing Service View & Export Terrain Analysis Spatial Analysis Desktop Viewshed Contour Proile - Edit Hydrology Viewshed Contour Proile Volume Estimates

57 Summary Serving Elevation data involves many details! Features in ArcGIS 10 ocused on improving eiciency and data management - Mosaic Dataset - Image Services via SOAP, REST, WMS, WCS, KML Best Practices worklows are under development - Emphasis on the Data Manager Increased LiDAR support Much more to come!

58 Please Fill Out Evaluation Forms Your comments help us meet your conerence needs each year. Thank you Any Questions?