Using 3 D Landscape Visualization to Simulate Forest

  • Slides: 36
Download presentation
Using 3 -D Landscape Visualization to Simulate Forest Management Impacts David J. Buckley Pacific

Using 3 -D Landscape Visualization to Simulate Forest Management Impacts David J. Buckley Pacific Meridian Resources Joseph K. Berry & Associates Craig Ulbright Software Developer

Landscape Structure Analysis/Visualization Web Site: www. innovativegis. com To find out more about Landscape

Landscape Structure Analysis/Visualization Web Site: www. innovativegis. com To find out more about Landscape Structure Analysis and Visualization, visit our Web Site. You will find a wide array of information ranging from papers and presentations to project overviews and case studies. (Berry)

Landscape Visualization (3 -D Rendering) • Integration of existing technologies ü Commercial GIS software

Landscape Visualization (3 -D Rendering) • Integration of existing technologies ü Commercial GIS software has limited capabilities for realistic 3 -D visualization ü Scientific visualization software is readily available but at individual plot/stand level only ü Virtual reality techniques are increasingly used in computer games and film making. 3 -D Object Rendering (USFS SVS, 6/97) • Non-traditional GIS techniques ü 3 -D object rendering (trees, soil, buildings) ü Map animation (2 -D and 3 -D) ü Linking video/photos to maps “Multimedia GIS” (Berry)

3 -D Visualization Approaches Sports. Tracker (Map. Trek, , 9/98) Image Draping -- is

3 -D Visualization Approaches Sports. Tracker (Map. Trek, , 9/98) Image Draping -- is an established technique in GIS. Draping a topographic or thematic map onto a 3 -D terrain surface is effective but relies on abstract colors, shading and symbols. Draping a satellite image, such as a digital orthophoto, results in good surface texture and can produce visualizations suitable for depicting landscape-scale vegetation patterns. However, the 3 -D nature of landscape objects are not depicted and severe “pixelization” occurs in the foreground. 3 -D Analyst (ESRI, 9/98) 10 m SPOT using Arc. TIN (ESRI, 6/95) (Berry)

3 -D Visualization Approaches (continued) Video Imaging -- is a computer technique that "cuts-and-pastes”

3 -D Visualization Approaches (continued) Video Imaging -- is a computer technique that "cuts-and-pastes” digital photographic images to represent changes on the landscape. This approach produces high quality visualization output, but it is very manually intensive, contains no direct geo-referencing to a GIS database and often suffers from the artistic/subjective nature of the creation process. (Imaging Systems Lab, Brian Orland, Director, University of Illinois, Dept. of Landscape Architecture) Left image is an actual photo of western spruce budworm damage in the Deschutes NF, Oregon. Right image depicts what the same place might look like 5 years after the dead/damaged trees are removed. The diseased trees are “manually” erased from the scene and background trees and shrubs are inserted “pasted” in their place. (Berry)

3 -D Visualization Approaches (continued) Geometric Video Imaging -- combines video-imaging techniques with geometric

3 -D Visualization Approaches (continued) Geometric Video Imaging -- combines video-imaging techniques with geometric registration, typically undertaken within a GIS. It's rarely used on a production basis, due to the difficulty in accurately geo-referencing the photographic video image with the 3 -D perspective framework (wire-frames) generated by the GIS system. Geometric Modeling -- Virtual Forest (Innovative Gis, 9/98) builds a 3 -D geometric representation of individual landscape features, such as trees, represented as geometric objects (simple cones or complex tree shapes). The individual objects are assembled to create a forest stand or landscape view depicting the perspective from a given viewpoint. 3 -D Rendering utilizes GIS inventories of tree types and “Virtual Reality” densities. Smart. Forest (Orland, 9/98) (Berry)

3 -D Rendering: Tree Design … 3 -D object design (trees) The Tree Designer

3 -D Rendering: Tree Design … 3 -D object design (trees) The Tree Designer is used to build custom 3 -D tree symbols that represent forest components, such as different species, multi-age trees, seasonal effects, shrubs, and snags 3 -D trees with several sun angles are created and stored as bit mapped images for polygon rendering (Berry)

Landscape Visualization (Rendering Technique) “Laying the Carpet” Step 1) 3 -D Terrain Surface Step

Landscape Visualization (Rendering Technique) “Laying the Carpet” Step 1) 3 -D Terrain Surface Step 2) Polygon Containers Step 3) Surface Texture Step 4) Tree Objects Step 5) Final Composition Step 6) Atmospheric Effects “Pouring the Trees” (Berry)

Landscape Visualization (Rendering Technique) Rendered Scene GIS d. B Tree Symbol %Maturity #Trees

Landscape Visualization (Rendering Technique) Rendered Scene GIS d. B Tree Symbol %Maturity #Trees

Landscape Visualization (GIS Rendering) (Berry)

Landscape Visualization (GIS Rendering) (Berry)

Landscape Visualization (GIS Rendering) (Berry)

Landscape Visualization (GIS Rendering) (Berry)

Landscape Visualization (GIS Rendering) (Berry)

Landscape Visualization (GIS Rendering) (Berry)

Visualizing Forest Conditions …changing the landscape’s carpet and objects Before Fire After Fire Summer

Visualizing Forest Conditions …changing the landscape’s carpet and objects Before Fire After Fire Summer Winter After Snowfall (Berry)

Landscape Visualization (Case Study 1) …with classified forest polygons Darker boundary lines identify forest

Landscape Visualization (Case Study 1) …with classified forest polygons Darker boundary lines identify forest parcels used for 3 -D rendered scenes ü Species ü Stocking ü Height/age ü Ground Cover (Berry)

Landscape Visualization (Case Study 1) This is a general image of the watershed surrounding

Landscape Visualization (Case Study 1) This is a general image of the watershed surrounding the primary lake showing harvest blocks. Texture Mapping is used to provide a realistic sky and lake surface as well as underlying colors for the blocks. Note the different species in the foreground stands (bottom left). This represents a multi-species stand. Target Observer (Berry)

Landscape Visualization (Case Study 1) This view provides a ridge viewpoint overlooking the lake

Landscape Visualization (Case Study 1) This view provides a ridge viewpoint overlooking the lake and adjoining harvest blocks. Note that there is little distinction in density of trees for the nonproductive stands that cap the ridge areas compared to stands classified as productive on the hillsides. A rock slide also appears in this view in the lower right. Target Observer (Berry)

Landscape Visualization (Case Study 1) Trees have been reduced by 50% and re-planted for

Landscape Visualization (Case Study 1) Trees have been reduced by 50% and re-planted for the non-productive stands that cap the ridges. Different textures have also been applied to the nonproductive stands and the harvest blocks to aid in their distinction. Note the different underlying colors for the blocks in the valley at the end of the lake and the non-productive, less dense stands on the higher elevations. Target Observer (Berry)

Landscape Visualization (Case Study 1) This image presents a 'boat' viewpoint from the middle

Landscape Visualization (Case Study 1) This image presents a 'boat' viewpoint from the middle of the lake. The viewpoint is directly above the primary lake looking down the valley towards the harvest blocks. Observer Target (Berry)

Landscape Visualization (Case Study 1) This image illustrates how fog and haze can be

Landscape Visualization (Case Study 1) This image illustrates how fog and haze can be used to add more realism to a forested visualization. Note the enhancement of the ridge in the foreground on the right side of the image. Observer Target (Berry)

Virtual Fly-By Animation provides an added dimension to 3 -D visualization… movement in space

Virtual Fly-By Animation provides an added dimension to 3 -D visualization… movement in space and time …rapidly sequencing a series of 3 -D renderings creates a sense of movement “Fly-By” (Berry)

Rendering “Fidelity” Data Richness determines the “fidelity” of a landscape rendering. Ground cover and

Rendering “Fidelity” Data Richness determines the “fidelity” of a landscape rendering. Ground cover and tree objects are assigned to polygons based on inferred forest conditions (limited data richness)… through detailed forest inventories (robust data richness) UPSHOT!!! … 3 -D rendering introduces a new element to “map accuracy” 1) Spatial Accuracy (positional), 2) Thematic Accuracy (classification) and 3) Rendering Fidelity (cartographic) (Berry)

3 -D Rendering Summary The integration of GIS and visualization technologies provides an operational

3 -D Rendering Summary The integration of GIS and visualization technologies provides an operational platform for natural resource professionals and general public to consider the aesthetic impacts of landscape alternatives. “point and click” data retrieval The striking realism of the rendered visualizations enhance communication of management plans in a more effective form. Future developments will extend these capabilities to a fully integrated view allowing interactive query of the 3 -D rendered scene. . . Point-and-Click for text/data summaries (Berry)

Landscape Visualization (Case Study 2) Analysis of topographic information, provincial forest inventory and Landsat

Landscape Visualization (Case Study 2) Analysis of topographic information, provincial forest inventory and Landsat imagery of the Redfish Creek watershed near Nelson, BC, was used to derive landscape characteristics. Band 5 Band 4 Band 3 (Berry)

Landscape Visualization (Case Study 2) Most Similar Neighbor analysis using canonical correlation was applied

Landscape Visualization (Case Study 2) Most Similar Neighbor analysis using canonical correlation was applied to find the best correspondence between the intensively inventoried subset of the landscape, about 20 -25% of stands, and a less intensively sampled classified remotely sensed image… (ESSA Technologies, Vancouver, BC) …the classified map was combined with the DEM and an older Aerial photo interpreted derived inventory, to generate an updated inventory (1999) for each forest stand. (Berry)

Landscape Visualization (Case Study 2) Study area landscape with draped texture map, hydrography is

Landscape Visualization (Case Study 2) Study area landscape with draped texture map, hydrography is in blue, roads in brown and primary forest stands in gold (3, 699 polygons). Key viewpoint locations and orientation are identified. (Pacific Meridian Resources, Fort Collins, CO) (Berry)

Landscape Visualization (Case Study 2) The new inventory was then modeled using the Prognosis

Landscape Visualization (Case Study 2) The new inventory was then modeled using the Prognosis Model… growing trees, predicting mortality, establishing regeneration, predicting non-tree vegetation development, performing management activities, and calculating tree volumes to simulate the landscape for 50 years under different management regimes. Viewpoint 2 Viewpoint 3 Initial Conditions 1999 Unmanaged 2049 Patch-Cutting 2049 Balanced Regime 2049 (Berry)

Viewpoint 3 , Initial Conditions 1999 (Berry)

Viewpoint 3 , Initial Conditions 1999 (Berry)

Viewpoint 3 , Unmanaged 2049 (Berry)

Viewpoint 3 , Unmanaged 2049 (Berry)

Viewpoint 3 , Patch-Cutting 2049 (Berry)

Viewpoint 3 , Patch-Cutting 2049 (Berry)

Viewpoint 3 – Balanced Regime 2049 (Berry)

Viewpoint 3 – Balanced Regime 2049 (Berry)

Example Project Using Advanced 3 -D Visualization Techniques to Assess Historical Forested Landscapes GIS

Example Project Using Advanced 3 -D Visualization Techniques to Assess Historical Forested Landscapes GIS Applications in Forestry and Resource Management

Landscape Visualization Summary GIS technology provides a new environment for more effectively communicating landscape

Landscape Visualization Summary GIS technology provides a new environment for more effectively communicating landscape characteristics and conditions-moves maps beyond “abstraction to realism. ” 3 -D Rendering enables users to visualize forest inventory data and simulate. . . “what could be. ” Video Mapping enables capture data and images within its actual spatial context. . . “what is. ” (Berry)

Landscape Visualization (Video Mapping) Documenting “real-world” conditions Multimedia GIS can be used to document

Landscape Visualization (Video Mapping) Documenting “real-world” conditions Multimedia GIS can be used to document actual conditions. Video Mapping georegisters… ü Streaming Video ü Still Images ü Audio ü Text ü Data …of existing landscape characteristics High Rendering Fidelity (Berry)

Landscape Visualization (Video Mapping) VMS 200 by Red Hen Systems Imagery organized by When

Landscape Visualization (Video Mapping) VMS 200 by Red Hen Systems Imagery organized by When and Where GPS coordinates are “stamped” directly on the video tape every second… …a database linking the GPS coordinates and tape position is automatically generated and used to construct a map of imagery locations. (Berry)

Going Beyond Landscape Visualization Landscape Structure Analysis provides information on the patterns and arrangements

Going Beyond Landscape Visualization Landscape Structure Analysis provides information on the patterns and arrangements of landscape features… “puzzle pieces. ” GIS Modeling provides a wealth of analytical tools that characterize spatial relationships. . . “spatial interaction and context. ” (Berry)

Unabashed Plug for Map-ematics Surface Modeling Spatial Analysis Geo. Statistics

Unabashed Plug for Map-ematics Surface Modeling Spatial Analysis Geo. Statistics