Wetlands Investigation Utilizing GIS and Remote Sensing Technology for Lucas County, Ohio: a hybrid analysis.

Transcription

1 Wetlands Investigation Utilizing GIS and Remote Sensing Technology for Lucas County, Ohio: a hybrid analysis. Update on current wetlands research in GISAG Nathan Torbick Spring 2003

2 Component One Remote Sensing The ERDAS Imagine Expert Classifier has two main elements; the Knowledge Engineer and the Knowledge Classifier. The Knowledge Engineer provides methodology for users with advanced information and experience to define variables, rules, and classifying interests to design a hierarchical decision tree and knowledge database. The Knowledge Classifier provides methodology to utilize the knowledge database created by the user and Engineer.

3 ERDAS Imagine Knowledge Engineer & Knowledge Classifier Previous attempts at classifying wetland types provides confirmed accurate training sites that can be utilized. Using the inquirer cursor function and signatures editor precise pixel values and signatures can be extracted for an AOI. With the hierarchical decision tree a hypothesis can be created with rules defining variables. The Knowledge Engineer feature allows the user to define nearly every aspect of the image. Hypothesis Rules Variables

4 2km coastal zone buffer

5 Radiometirc Enhancement Atmospheric Haze Reduction For multi-spectral images, this method is based on the Tasseled Cap transformation which yields a component that correlates with haze. This component is removed and the image is transformed back into RGB space. For panchromatic images, an inverse point spread convolution is used.

6 Histograms of Landsat 7 ETM reflectance responses. Left graph displays single image. Right graph displays spectral response of a multitemporal stacked image with an applied tasselcapped based algorithm for atmospheric haze enhancement.

7

8 Pixel value/histogram examination

9 Knowledge database output over Landsat image

10 Knowledge database output over Lucas aerial photo. Coastal Green = Wet Forest Yellow = Wet Prairie

11 Example of classified Landsat 7 ETM+ output from ERDAS. Ottawa Park off Brancroft St. across from UT. Blue = Wet Forest Red = Wet Prairie

12 Component Two. GIS

13 Soils DEM Drift thickness Bedrock/Geology Slope Layered LayeredMa p Map Matrix Matrix Landsat ETM classification Floodplains Water Table Hydrology Ground Water Wetlands Fig. 1. Conceptual framework for wetlands cartographic model.

14 Data Collection Flood data was complied from Ohio Department of Natural Resources. The flood data layer is a combination of the 100 & 500 year floodplains and flood hazard areas in Lucas County. The ODNR GIS also provided soils data.

15 Some wetland characteristics and parameters required large amounts of time and analysis for collection and processing. Water table depth, drift depth, and bedrock geology data was acquired from published hydrogeology reports. Maps required scanning into digital format, georeferencing, and digitizing.

16 Watershed Drainage Network. Channel networks with arbitrary drainage or resolution can be extracted from digital elevation data (Tarboton 1991). This method is based on elevation gradients, flow accumulation, and drainage networks. Using Arc/INFO a flow-order-accumulation network can be designed. This takes substantial experimenting to capture the desired results (1.53km 2).

17 GPS integration. Wet Prairie near Kitty Todd reserve.

18 GPS Integration.

19 Soils DEM Drift thickness Bedrock/Geology Slope Layered LayeredMa p Map Matrix Matrix Landsat ETM classification Floodplains Water Table Hydrology Ground Water Wetlands Fig. 1. Conceptual framework for wetlands cartographic model.

20 Model Simulation and Analysis. A rating system has been/is being developed and tested for each model parameter. Within each coverage, value fields have been added to attribute tables and reclassified for model input. The amount of weight, percentage of input, and strength of variable for input to the model can be manipulated for desired results. This is the main human interaction/element in the project. Different scenarios with changes in parameter strength and methods can be run outputting different end results ranging from regression analysis to constraint mapping to weighted variables.

21 Scenario One. Percentage of Input - Strength Scenario One Output Table

22 q Theory: Higher values reflect wetland characteristics Matrix output for wetland model.

23 Future Directions CAUV & Additional season scenes Ground truthing/accuracy Assessment Summer REU Accessibility of Data Visual Basic Code of model manipulation and parameter adjustments