Processing Remote Sensing Data with Python

Processing Remote Sensing Data with PythonCONTENTS1 Acknowledgements2 Introduction2.1Gcosctup2.2 Improving this Documentation3 Brief Overview of Coordinate Systems and Map projections3.1 Coordinate Systems557783.2 Earth's Shape and Datums3.3 Map projections84 Getting Started with Ocean Color and Bathymetric Data4.1 Data Formats4.2 Data Properties124.3 Focus datasets134.4 Previewing Data5 Getting Started with Python215.1 Science and Python215.2Troubleshooting215.3 Important things to consider225.4 Setting up your development environment226 Importing netcDF and HDF Data with Python256.1 netCDF47 Working with Projections Using Python297.1 pyproj......298 Gridding and Resampling Data with Python318.1 Deciding what to do318.2 Gridding,,318.3 Re-sampling to Grid329 Plotting with Python359. 1 Matplotlib and Basemap3510 Writing Geospatial Data to Files3910.1 Write file function3911 Geosetup-textdata4312 Geosetup- Pathfinder4513 Geosetup-CORTAD4914 Geosetup-Glob Colour314.1 Importing Mapped l3m Data14.2 Importing isin grid L3b Data515 Geosetup-gebco5716 Geosetup- Main Program17 Appendix: Temporal Coverage of Ocean Color Satellite Missions6717.1 MissionsBibliography69CHAPTERONEACKNOWLEDGEMENTSFirst, I would like to thank the American Scandinavian Foundation in their support of mywork. I would also like to thank guorun marteinsdottir for her coordination and advising,and Thorvaldur Gunlaugsson and Mette Mauritzen for their advising on my project for whichthis coding is intended to support. Lastly, I am very appreciative toward Gisli Vikingsson andHafrannsoknastofnun for arranging a place for me to workProcessing Remote Sensing Data with Python Documentation, Release 1Chapter 1. AcknowledgementsCHAPTERTWOINTRODUCTIONWith public access available for numerous satellite imaging products, modelling in atmosphericand oceanographic applications has become increasingly more prevalentThough there are numerous tools available for geospatial development, their use is more commonly applied towards mapping applications. With this being the case, there are a number ofvaluable texts for using these tools in such mapping applications [11][1]; though, documenta-tion for processing of remote sensing datasets is limited to brief contributions on personal blogsor manual pages and tutorials specific to one library or dataset. Python programming methodsfor performing such tasks will be focused on here, collecting various code and informationfrom these text, blogs, etc. and presenting original code that may be employed in scripts toperform commonly required tasks in processing remote sensing dataFor a general place to get started with geospatial work with Python, Erik Westra's"PythonGeospatial Development"is an excellent resource, and it will provide an excellent overview ofgeospatial work with python before looking at specific remote sensing datasets2.1 GeosetupThe tools presented here are collected into a Python program repository (currently under development)that will produce standardized gridded satellite data files that may be used as input forvarious models. This repository is publicy available on the collaborative coding site GithubLinktorepositoryongIthub:https:/github.com/ryanjdillon/geosetupOr yo' ou may download the code as a zip-archiveIf you are interested in collaborating on this cffort, plcasc fccl encouraged to fork thc repositoryon github and offer your contributions2.2 Improving this DocumentationThere are many improvements that can always be made in either the code shown here or theexplanation of a particular approach to something. If you find something broken, or stuck withan inadequate explanation, feel free to contact me so that I might correct itProcessing Remote Sensing Data with Python Documentation, Release 1Thanks,Ryanryanjanesdillon@gmail.comChapter 2. IntroductionCHAPTERTHREEBRIEF OVERVIEW OF COORDINATESYSTEMS AND MAP PROJECTIONSBefore working with georeferenced datasets, it is important to understand how things can bereferenced geographicallyPositions on earths surface can be represented using different systems, the most accurate ofthese being geodetic positions determined by coordinates from a particular geographic coordi-nate system, or geodetic systemIn addition to coordinate systems, the assumed shape of the earth will affect how coordinatepositions are translated to earths surface. Geodetic Datums are mathematical definitions offor the shape of the earth, which in turn defines a geodetic system3.1 Coordinate SystemsCoordinate systems are categorized into two groups, both of which being commonly used inrepresenting remote sensing dataunprojected coordinate systems these are 3-dimensional coordinate systems, such aslatitude and longitude (referred to as geographic coordinate system in common GIssoftware)projected coordinate systems these are 2-dimensional, there are many all of which having an advantage over another for a particular useEither system may be used depending on the data format(e. g netCDF HDF containing arraysof unprojected data vs. GeoTIFF images containing projected data). When working withmultiple datasets, or doing some processing of a dataset, you may need to change betweendifferent coordinate systemsMost often data will coincide with a latitude and longitude(i.e. unprojected ). If it is desiredto interpolate or re-sample this data, transforming the data to a projected system allows forsimpler and a greater variety of methods to be used