dlk253 PUI2016 Extra Credit Proposal



UPDATE:
This proposal was updated to reflect the results seen in the report located here.
After speaking with the Prof. Fedhere, the analysis shifted from histograms to plotting geographic features onto the raster image. 
The tiling of the images and hillside were completed as originally proposed. 


PUI2016 Extra Credit Project Proposal
Anthropogenic Impacts found the Bedrock Layer<Dana Karwas, dlk253, dlk253>


Problem Description: Coastal Urban ecosystems are under the constant pressure of natural and man made forces. How much is the bedrock layer effected by urban coastal ecosystems? By identifying patterns at the bedrock layer is it possible to identify urban coastal areas through their bedrock profile? Can an algorithm to measure anthropogenic impacts on urban coastal ecosystems be established by applying visual synthesis and analysis techniques to bedrock models? What can the bedrock tell us about the current state of our coastal cities? Can a metric for human impact be established by looking at the shape and topographic details of the bedrock? 

Data:  

The dataset this that  is available and suitable is the Earth 2014 arcmin global topography and relief models from Curtin University. The data includes a global bedrock only layer which is what I would like to start with. It is available as gridded data and degree‐10,800 spherical harmonic. The bedrock (BED) includes Earth`s relief without water and ice masses.

This data was found it in the paper linked below with accompanying data gateway.  

Data  Contact: c.hirth@curtin.edu.au

This data is suitable for my questions because it has an isolated bedrock layer for the entire globe. The analysis will be made on three coastal urban cities in the US (New York City, Los Angeles, and New Orleans). I will have to pay close attention to land and ocean stitching and may need to find additional data to fill in data gaps in resolution if needed.  I will also need to play close attention to the coordinate system transformation for my datasets.  I will look for topographic anomalies by using imaging processing techniques on the shape files.  I will establish a search criteria through image processing (histogram matching/analysis) for "man-made" interventions - ultimately leading to machine learning (this is very ambitious, and would be happy if I could just begin to compare a few histograms).

Other data of interest: 
Earth
1- ETOPO1 (1 arc minute)
2- SRTM30_PLUS (0.5 arc minute ~ 900 meters) and SRTM15_PLUS (0.25 arc minute ~ 450 meters)

Mars
The MOLA Mission Experiment Gridded Data Records (MEGDRs) are global topographic maps of Mars 


Analysis 

Image processing techniques such as histogram matching could be used as a way to compare the datasets. Finding patterns in the histograms would be one way to begin identifying the impacts. 

References

Technical References:


Theoretical References:



Deliverable: 

The expected deliverable would be an algorithm that can stitch together topographic bedrock data of ANY planetary body, href="http://astrogeology.usgs.gov/search/map/Mars/GlobalSurveyor/MOLA/Mars_MGS_MOLA_DEM_mosaic_global_463m" target="_blank">such href="http://astrogeology.usgs.gov/search/map/Mars/GlobalSurveyor/MOLA/Mars_MGS_MOLA_DEM_mosaic_global_463m">such  as mars-- and search for human impact in that dataset.  This algorithm can be used by agencies and students to search for "unnatural impacts". This will be interesting, I think, when the impact includes errors from the sensing device - such as those discussed in the mars MOLA dataset and impacts created from human  (or other) intervention.