Env Remote Sensing Lab
Env Remote Sensing Lab ERS 186L
Popular in Course
Popular in Environmental Resource Science
Dan Skiles IV
verified elite notetaker
This 46 page Class Notes was uploaded by Dan Skiles IV on Tuesday September 8, 2015. The Class Notes belongs to ERS 186L at University of California - Davis taught by Staff in Fall. Since its upload, it has received 92 views. For similar materials see /class/187672/ers-186l-university-of-california-davis in Environmental Resource Science at University of California - Davis.
Reviews for Env Remote Sensing Lab
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 09/08/15
Open Space How Urbanization and Agriculture impact the View Shed around the City of Davis Introduction 0 The development of land has a dramatic impact on the scenic aspects of the area and greatly affect the feel of the area Urbanization The existence and growth of the urban landscape blocks the View of the outlying landscape and detracts from its scenic value Agriculture 0 Agriculture also impacts the View Shed It especially detracts from the scenic Views after harvest and during fallow periods Problem Has the View Shed Around Davis Changed over the years If it has Changed What has Changed and by how much Images Acquired Images glefumiaesumdeduindeXshtml 1998 Landsat TM and 2000 Landsat ETM Crop images to same size and extent Smaller images were faster to process Calibration Empirical Line A total of 50 points were used when nding the light and dark pixels shared by the 1998 and 2000 images They were then used to create an Empirical Line image using the 1998 image as a reference Classi cation Determine Classes Combine Classes Trees View Shed Grasses 39 Trees Water Grasses Agriculture ygillture Soil Open Urban Detractor Open Urban Soil Classi cation Create Regions of Interest for each classi cation consisting of several pixels for each category Use these ROI s to classify the rest of the image using the Maximum Likelihood function in ENVI Confusion To determine the accuracy of the classi ed image a confusion matrix was created The classi cation was found to be 84 accurate Once the classes were combined to View Shed and Detractor the class image was found to be 9189 accurate Classi cation The same classi cation parameters were then used to create the 2000 class image Another confusion matrix was run with all classes and with the combined classes Accuracy was found to be 6434 for all classes and 8278 with the combined classes Analysis Since both images were classi ed using the same parameters it was possibly to compute the area for each class and its percentage of the total area The images were then compared to determine any changes between them Result It was found that the View Shed around the Davis area decreased by 83725 hectares over the 12 year period This corresponds to a 722 change over 12 years Or about 060 annually Conclusion Although Davis has expanded over the years the Change was relatively small given the extent of the image Vegetation Changes in Davis between 1989 and 2000 Ginger Kennedy Question Did Davis Expand Hypothesis Yes 1989 Image 2000 Image 1989 Empirical Line 1989 Davis Image 2000 Davis Image 1989 NDVI Image Bright Pixels More green vegetation Dark Pixels Less green vegetation 2000 NDVI Image Bright Pixels More green vegetation Dark Pixels Less green vegetation Compute Difference Map Color Shades Red Positive change Gray No change Blue Negative change Band Math Bright Values Increase in green vegetation Dark Values Decrease in green vegetation Statistics Report and Histogram Min Max Mean Stdev 1042177 0950562 0034511 0234898 Band Math with Density Slice Red Decrease in green vegetation Green No change in green vegetation Blue Increase in green vegetation Results Decrease in green vegetation 9083 pixels 73777 hectares No change in green vegetation 47399 pixels 384998 hectares Increase in green vegetation 7393 pixels 60050 hectares Davis i Expanding EWQ Mapping I nvasive Species Kudzu top and Lovegrass bottom are two invasive species at FT Benning GA 0 Before classification comes lots and lots of processing Assembling AVIRIS Fl39ght Lines Creating HeaderFiIes from each individual scene Each Scenes Header file can be created from readme files from the raw image data from JPL n ghiiins n spam its m Beg 60W 3 i Emmi 4E ZtW S ZSE39EZV i n ghiiins t2 Beg E NJZ team imam 72w n mine 4 Bay 39 3M ENW Fi ghiimz ii Beg 39 13 Eini fi ghiiins tf 39t WUW iBZBN V323 Fi ghiiine ttit Begin 3435 EEVHZ i7 YEN E 1 v H J I Aishaquot MSW End iiUEW EBZU i FmBentinu I i N ngmimeen 39 Eit39t BeginEASEEIZW mm Emi i 2Ei Creating Flight Line Mosaic The Map Info for each header file must be calculated so that the flight line is in correct or This can be done manually by editing each individual header file with the proper coordinates It was easier to assemble scenes via pixel coordinates relative to the first scene s geographic coordinates Beginning Coordinates of first scene top center minus 12 of e coordinates of pixel 1 1 before rotation More Image Processing Atmospheric Noise Removal Correction with ACORN Minimum Noise Dark Object Fraction Subtraction and Empirical Line Calibration Continuum Removal ACORN atmospheric correction now as opposed to atmospheric correction later Spectral Calibration Gain and Offset Files must be prepared beforehand from JPL files Information is then converted to excel or ASCII format After clicking the RUN button for ACORN you will then become an expert at interpreting ERROR messages Imagery and Zprofile of a Kudzu ROI before ACORN ml Pro le Image is still in radiance Imagery and Zprofile of a Kudzu ROI after ACORN Type 12 amp3 artifact suppression Note the reduction of the water absorption bands Imagery is now converted to reflectance and is ready for further processing Dark Object Subtraction before ACORN Dark Subtraction was applied to correct for atmospheric scattering conditions This method was applied to both before and after ACORN images to see the differences Dark Object after ACORN Oops Need to mask after ACORN to mask out the invalid data L E liranxwwuuh After Masking out invalid data Image is now ready for classification Essentially the same image as the ACORN output image Images are ready for Noise Reduction Minimum Noise Fraction removed excess noise from the bands and reduced the dataset Continuum Removal was performed to normalize reflectance spectra to allow comparison of individual absorption features Continuum Removal The Continuum Removal Function is designed to help smooth out the data Spectral Endmembers Developing spectral endmembers from field data Exporting field data into the Spectral Analysis Management system Export Data into ENVl s spectral endmember library Regions of Interest The next step is These shapefiles are transform the field then exported from spectral data into GPS pathfinder into shapefiles an ENVI shapefile Using GPS pathfinder format identify your ROls The shapefiles can from an ASD data file then be laid over the images as vectors Unsupervised Classification Isodata Classification 30 classes chosen Compare ROls of each classification Classification outcomes were visibly different Unsupervised Classification K means Classification Same parameters as Isodata classification ROI comparison shows visibly different data