Methods

For this project the core of our research will rely on utilizing the Envi Classic platform where we will use Landsat 8 satellite data derived from the USGS website. With this data, we will employ standard remote sensing techniques in an attempt to determine the most viable corridor placements for Rim of the Valley wildlife. Landsat 8 data was acquired through glovis.usgs.gov for the month of November in the year 2013 to show variance in the vegetative cover via the Normalized Difference Vegetative Index (NDVI) and to compare the ground cover with thermal imagery.

For the NDVI imagery we classified the data with a density slice to help sort out different vegetation types and to detect anomalies such as burn areas and irrigation. This density slice was classified by intense high resolution imagery scrutiny and prior knowledge of the areas from living in the region. The classifications were kept limited to six major classes so as to minimize clutter due to large amounts of scatter present throughout the region by heavily vegetated urban areas like the San Fernando and Conejo valleys. To improve confidence in the density slice classifications were assembled from an East to West profile that bisects the Las Virgenes Reservoir and passes just south of Lake Sherwood through the country club providing most of the classified areas in relatively close proximity. From this spectral profile we were able to narrow down the classes as follows: 

DENSITY SLICE VALUES:
-0.1300 to -0.0380 Body of Water
-0.0380 to -0.0240 Shoreline / Shallow Water 
-0.0240 to 0.0700 Burn Area / Low Turbidity / Urban
0.0700 to 0.1240 Chaparral Grassland - Shrubland / Urban Vegetation
0.1240 to 0.2090 Chaparral Dry Forest / Urban Tree Cover
0.2090 to 0.2900 Chaparral Oak-Sycamore Woodland / Riparian / Urban Tree Cover
0.2900 to 0.5487 Lawn / Golf Course / Agriculture

We will continue to analyze the NDVI data with ARCMAP 10.2 by layering the NDVI data of individual study areas over color high resolution satellite imagery. This is done, not only to verify the density slice classification, but to identify large or small patches of habitat that are adjacent to the study area. This was done by exporting the individual NDVI density slices of regions of interest into .tiff files which can  be easily used by the ArcMAP 10.2 program. Then we used the “Satellite Imagery” base-layer as    This provides much needed analysis of habitat connectivity by direct color image observation as we are able to make connections between NDVI data directly over the color, high resolution satellite imagery. The accuracy of the density slice can be observed by the locked lines of the freeway systems, the obvious bodies of water and the stark and apparent burn area in the eastern portion of the Outline image, or in the Conejo Grade Study Area,  represented as the Springs Fire burn area.

To assess the terrain and relief of the individual study areas we decided to do a side by side comparison of NDVI and thermal data draped over a three dimensional surface derived with Shuttle Radar Topography Mission (SRTM) data. To do this we created regions of interest focused on the individual study sites with the ArcMAP 10.2 program. These regions of interest were drawn from personal knowledge of the study areas over roadmap imagery for spatial reference and were given a wide berth around the potential corridor sites in order to have as much data as possible without cluttering the images. With the spatial subsets of the study areas created as regions of interest in shapefile format, it was then a simple task of subsetting the NDVI file, the SRTM file and the thermal bandwidths to the regions of interest. For the NDVI 3D image, a snapshot of the density slice was lain over the SRTM data for each individual region. For the thermal imagery we began by creating a large image for direct comparison to the NDVI Outline image and decided on the RGB band designations being:  

RED = Band 11 Thermal Infrared (TIRS) 2, 11.5-12.51 micrometers.  

GREEN = Band 10 Thermal Infrared (TIRS) 1, 10.6 - 11.19 Micrometers.  

BLUE = Band 5 - Near Infrared (NIR), 0.85 - 0.88 micrometers.

This created an easy to read thermal map with blues representing cooler colors and yellows representing warmer colors as is displayed in the legend. Mountain tops contrast sharply as dark, cold blues whereas the urban areas give off dark, warm yellows; of note is the very warm and obvious Springs Fire burn area on the bottom left side of the image. With this base outline image of thermal data we then repeated the previous steps taken with the NDVI image and created five separate images of thermal data corresponding to the study areas to then be used as a surface for the SRTM data. With the ten 3D images (5 NDVI, 5 Thermal) ready to be captured, we then aligned the corresponding 3D images of the individual study areas so that they showed the most relevant positions for assessing the data, as well as for useful visual references.

Finally, actual corridor placement suggestions will be shown over low elevation high resolution Google Earth imagery to improve accuracy of site selection. These final placement suggestions will be based on the combination of NDVI imagery, Thermal imagery, terrain data and high resolution color satellite imagery and should show the optimum placement for future wildlife corridors in the Rim of the Valley