GIS – Whirlwind TourGIS – Whirlwind Tour(In 25 slides)(In 25 slides)

What, Why, HowWhat, Why, How

GISGIS► Geographical Information ScienceGeographical Information Science

Not just computerised mapsNot just computerised maps

► Data Capture ('EO' – though often = just physical/RS)Data Capture ('EO' – though often = just physical/RS) Survey: GPS, EDM, LaserscannerSurvey: GPS, EDM, Laserscanner RS: Aerial/Satellite, but also other sensors, Sensor NetworksRS: Aerial/Satellite, but also other sensors, Sensor Networks Primary Data, Secondary Data (verification techniques/theory/PAI)Primary Data, Secondary Data (verification techniques/theory/PAI)

► Analysis, 2D Map – Cartography, also whole field of Analysis, 2D Map – Cartography, also whole field of (Geo)Visualisation, (incl. 3D)(Geo)Visualisation, (incl. 3D)

► Visualisation can also permit further analysisVisualisation can also permit further analysis Exploratory (Spatial) Data Analysis – EDA/ESDAExploratory (Spatial) Data Analysis – EDA/ESDA

GIS – Three or Four KindsGIS – Three or Four Kinds

►Desktop Application / Full PackageDesktop Application / Full Package

►Web Mapping / Feature Server / Server Web Mapping / Feature Server / Server GISGIS

►Web Browser with GIS Tools / Thick ClientWeb Browser with GIS Tools / Thick Client

►Apps, Mashups, APIs – Distributed GISApps, Mashups, APIs – Distributed GIS

Spatial PhenomenaSpatial Phenomena

► Land Use – Urban, Rural, Building TypesLand Use – Urban, Rural, Building Types

► Flood Risk, Water Transport, Soil TypeFlood Risk, Water Transport, Soil Type

► Topography – Elevation, Slope, AspectTopography – Elevation, Slope, Aspect How does topog’ affect ‘occurrence’ in landscapeHow does topog’ affect ‘occurrence’ in landscape

► People – Travel to work, shop, emergency People – Travel to work, shop, emergency servicesservices

ModellingModelling► Conceptual ModelsConceptual Models

To understand world, predict conditions at locations in time and/or To understand world, predict conditions at locations in time and/or spacespace

► Mathematical ModelsMathematical Models Numerical models where formalised - some idealised, some less soNumerical models where formalised - some idealised, some less so

► Data ModelsData Models Structure and flow of information in time and spaceStructure and flow of information in time and space

► Spatial Data – often (not always) represented in mapsSpatial Data – often (not always) represented in maps (Lots of) Data with spatial component, some attempts to address time (Lots of) Data with spatial component, some attempts to address time

tootoo

► Computerised Spatial Data -> Quick Computerised Spatial Data -> Quick SpatialSpatial Analysis over wide Analysis over wide extentextent GIS – Geographical Information Science (and/or Systems)GIS – Geographical Information Science (and/or Systems)

GIS History / SoftwareGIS History / Software► Geography Techniques (by hand) pre 1960s: John Snow, Minard’s Map Geography Techniques (by hand) pre 1960s: John Snow, Minard’s Map

(Napoleon)(Napoleon)

► Forestry – Canada (+E Africa) - CGISForestry – Canada (+E Africa) - CGIS First GIS – Roger Tomlinson 1960+, operational from 1971+First GIS – Roger Tomlinson 1960+, operational from 1971+

► USA – Government Organisations: USGS, US Forest Serv, others incl. CIAUSA – Government Organisations: USGS, US Forest Serv, others incl. CIA

► AcademiaAcademia EdinburghEdinburgh – GIMMS 1970+ (Sold from 1973), MSc GIS 1985+ – GIMMS 1970+ (Sold from 1973), MSc GIS 1985+ Harvard – Computer Graphics and Spatial Analysis Lab 1965Harvard – Computer Graphics and Spatial Analysis Lab 1965

► ESRI 1969 Env. Consultancy – Arc/Info 1982 -> ArcView Desktop 1995 -> ArcGIS ESRI 1969 Env. Consultancy – Arc/Info 1982 -> ArcView Desktop 1995 -> ArcGIS 19991999

► Physics/Space (Moon landings) later CAD/Utilities – LaserScan/Intergraph 1969Physics/Space (Moon landings) later CAD/Utilities – LaserScan/Intergraph 1969

► Demographics/Consultancy – MapInfo 1986Demographics/Consultancy – MapInfo 1986

► OpenSource – GRASS, Quantum GIS (QGIS), gvSIG, … link to DBMSOpenSource – GRASS, Quantum GIS (QGIS), gvSIG, … link to DBMS

► Web GIS – WMS, WFS, Google Maps, Google Earth, OGC, OpenStreetMapWeb GIS – WMS, WFS, Google Maps, Google Earth, OGC, OpenStreetMap

Data TypesData Types

►Vector – Discrete Entities Vector – Discrete Entities withinwithin space space PointsPoints LinesLines PolygonsPolygons

►Raster – Contin’s Field/Surface Raster – Contin’s Field/Surface acrossacross space space ElevationElevation pHpH Growth Pot’l as secondary data based on aboveGrowth Pot’l as secondary data based on above

AttributesAttributes

► Vector – Multiple Attributes (Properties)Vector – Multiple Attributes (Properties) Attributes are of each feature (point, line, poly)Attributes are of each feature (point, line, poly)

► Raster – Single Attribute (Value) e.g. pHRaster – Single Attribute (Value) e.g. pH Each cell has a different value of this attributeEach cell has a different value of this attribute

BUT! Can also have in turn BUT! Can also have in turn Value AttributesValue Attributes e.g. e.g. 1 = Acid, 7 = Neutral, 14 = Alkaline1 = Acid, 7 = Neutral, 14 = Alkaline

BUT! Again only one per value!BUT! Again only one per value!

Model Framework to use – 1?Model Framework to use – 1?

►Q. For mapping HGVs across Europe?Q. For mapping HGVs across Europe?

Ans: Lines – A Linear Network (Vector)Ans: Lines – A Linear Network (Vector)

Lorries constrained to linear road networkLorries constrained to linear road network

Each road can have multiple attributes: Each road can have multiple attributes: speed limit, length, width, number of speed limit, length, width, number of laneslanes

Model Framework to use – 2Model Framework to use – 2

►Q. To model flow/drainage in moorland?Q. To model flow/drainage in moorland?

Ans: Raster Grid – A continuous surfaceAns: Raster Grid – A continuous surface

Each cell can have a flow directionEach cell can have a flow direction

Need multiple spatially co-incident grids to Need multiple spatially co-incident grids to combine in order to achieve end result combine in order to achieve end result (answer)(answer)

Spatial Co-incidence – Map Spatial Co-incidence – Map LayersLayers

► Combination of spatial and aspatial (often Combination of spatial and aspatial (often numerical) manipulation of datanumerical) manipulation of data

► Grids lie on top of each other. Co-incident cells can Grids lie on top of each other. Co-incident cells can then be combined numerically to give result.then be combined numerically to give result.

► GIS all about combining info from different GIS all about combining info from different LayersLayers

► Layers form a stack – but usually only in model – Layers form a stack – but usually only in model – multiple measures found in same x,y,z (cell) locationmultiple measures found in same x,y,z (cell) location E.g. elevation, pH, salinity – each of these in different grid E.g. elevation, pH, salinity – each of these in different grid

layerlayer

Overlay – Attribute TransferOverlay – Attribute Transfer

► Can convert between raster + vector but Can convert between raster + vector but limited and tend to be treated in isolation but limited and tend to be treated in isolation but can be viewed together easilycan be viewed together easily

► Can however easily combine vector layers – Can however easily combine vector layers – mathematical combination of geometry – mathematical combination of geometry – easy to cut-up and intersect => Vector easy to cut-up and intersect => Vector OverlayOverlay

► Vector Overlay all about Attribute TransferVector Overlay all about Attribute Transfer

Overlay – Point in PolygonOverlay – Point in Polygon

► Which district has the Which district has the most towns?most towns? Count the number of town Count the number of town

points in each district polypoints in each district poly

► In which district does this In which district does this town lie?town lie? Attribute (verb) each town Attribute (verb) each town

with the name of the with the name of the district polygon in which it district polygon in which it fallsfalls

► Points 'lie on top' of solid Points 'lie on top' of solid coloured polys in our coloured polys in our stackstack

Overlay – Polygon OverlayOverlay – Polygon Overlay

►Degree of overlap between different Degree of overlap between different districts/zones/catchment areasdistricts/zones/catchment areas

►E.g. E.g. EraseErase SSSI polygons from potential Golf SSSI polygons from potential Golf Course polygons (unless you are a Trump)Course polygons (unless you are a Trump)

► Intersect pollution zones with population Intersect pollution zones with population zones (> 10,000 pop) to get danger zones!zones (> 10,000 pop) to get danger zones!

Co-ord’te Reference Systems Co-ord’te Reference Systems (CRS) – (CRS) –

Map Projections, DatumsMap Projections, Datums► Spatial Data can be measured/located in:Spatial Data can be measured/located in:

Angular Units – Lat, Long, e.g. 56Angular Units – Lat, Long, e.g. 56º23'4'' (dms), 56.38º (dd)º23'4'' (dms), 56.38º (dd) Linear Units – Flat Grid-based: Easting, Northing, e.g. metres, ftLinear Units – Flat Grid-based: Easting, Northing, e.g. metres, ft

► Spherical (Angular) = 'Geographic' CRS (unprojected)Spherical (Angular) = 'Geographic' CRS (unprojected)

► Flat Planar (Linear) = 'Projected' CRS, e.g. BNG, UTMFlat Planar (Linear) = 'Projected' CRS, e.g. BNG, UTM

► All CRS based on a reference datum – a model of the Earth’s All CRS based on a reference datum – a model of the Earth’s surface/shape. This MUST be correctly defined, for any later surface/shape. This MUST be correctly defined, for any later projection (curved to flat) to WORK correctly.projection (curved to flat) to WORK correctly.

► If collecting GPS data in Britain, we want to end up in BNG but If collecting GPS data in Britain, we want to end up in BNG but MUST define source data as WGS84 datum / geographic sys MUST define source data as WGS84 datum / geographic sys as THAT is what the GPS uses. Once source data defined we as THAT is what the GPS uses. Once source data defined we can project to BNG.can project to BNG.

Simple GIS – Google EarthSimple GIS – Google Earth► Last point less of an issue if using Simple GIS – Google Earth – uses Last point less of an issue if using Simple GIS – Google Earth – uses

WGS84 lat long; loads in KML files – now often saved by GPS software WGS84 lat long; loads in KML files – now often saved by GPS software directly. E.g. GPS Utility. Or just write raw KML or use converter prog.directly. E.g. GPS Utility. Or just write raw KML or use converter prog.

► Simple annotation / measurement tools etc. but also clever features, Simple annotation / measurement tools etc. but also clever features, e.g. timestamp allows animation/viewing change through time e.g. timestamp allows animation/viewing change through time

► Beware – Google Beware – Google MapsMaps uses its own Mercator projection but you can uses its own Mercator projection but you can link to KML URL in Google Mapslink to KML URL in Google Maps

► Can make 'mashups': Google Maps, JavaScript, WMS, Scanned MapsCan make 'mashups': Google Maps, JavaScript, WMS, Scanned Maps

NO pricey GIS package requiredNO pricey GIS package required

Industry Standard - ArcGISIndustry Standard - ArcGIS► ArcGIS: ArcMap – 2D, ArcScene 3D, ArcCatalog, othersArcGIS: ArcMap – 2D, ArcScene 3D, ArcCatalog, others

► Relatively easy to get started, though can at times be Relatively easy to get started, though can at times be overwhelming! Some similarity to Word/Excel in overwhelming! Some similarity to Word/Excel in structurestructure

► ArcToolbox now primary interface to functionality (or ArcToolbox now primary interface to functionality (or command line) though various toolbars (drop-down command line) though various toolbars (drop-down menus) toomenus) too

► Beware – from v10 Arc tries to save everything to a Beware – from v10 Arc tries to save everything to a default geodatabase in user’s home path. In UoE, default geodatabase in user’s home path. In UoE, home path is M: and for undergrads this may still be home path is M: and for undergrads this may still be quite small. Thus must keep some space on M: even if quite small. Thus must keep some space on M: even if other space availableother space available

File FormatsFile Formats► Shapefile – Actually a set of 3-6 files (min 3)Shapefile – Actually a set of 3-6 files (min 3)

Prob one of most widely used file type – though closed, proprietary Prob one of most widely used file type – though closed, proprietary formatformat

Myfile.shp (Myfile.shp (geomgeom), Myfile.shx, Myfile.dbf (), Myfile.shx, Myfile.dbf (attrsattrs), .prj, ...), .prj, ... Move each file, or .zip all together at OS; just move the .shp in Move each file, or .zip all together at OS; just move the .shp in

ArcCatalogArcCatalog

► GeodatabaseGeodatabase Single file at OS level – neatly holds all vectors & rastersSingle file at OS level – neatly holds all vectors & rasters File Geodatabase now 1Tb (and there are ways to get up to 256 Tb!)File Geodatabase now 1Tb (and there are ways to get up to 256 Tb!) In time will likely be the only thing to use (other GIS still use shp for In time will likely be the only thing to use (other GIS still use shp for

now)now)

► Coverage (vector); Grid (raster)Coverage (vector); Grid (raster) Older formats you may need to know aboutOlder formats you may need to know about Hybrid structure of one folder mygrid and a shared info folder – shared Hybrid structure of one folder mygrid and a shared info folder – shared

between ALL coverages and grids in a containing folderbetween ALL coverages and grids in a containing folder Move ALL of these at OS – or better still use ArcCatalog only!Move ALL of these at OS – or better still use ArcCatalog only!

Data Storage TheoryData Storage Theory

► Hybrid Arc/Info model based on storing correct Hybrid Arc/Info model based on storing correct typetype of data in best place for thatof data in best place for that

► Data can be re-joined when requiredData can be re-joined when required

► Same principle means store only relevant info in a Same principle means store only relevant info in a table of particular feature types and join these at table of particular feature types and join these at query/display time – also use key tables & numbers to query/display time – also use key tables & numbers to reduce data volsreduce data vols

► Relational DBs good for this and can store spatial Relational DBs good for this and can store spatial data – Many offer spatial extensions with spatial data – Many offer spatial extensions with spatial analysis functionsanalysis functions

Database StorageDatabase Storage

►Can connect GIS to RDBMS for:Can connect GIS to RDBMS for: Better queryingBetter querying Robust storageRobust storage Multi-user access and sophisticated control Multi-user access and sophisticated control

(only one user edits electoral district at a (only one user edits electoral district at a time!)time!)

►Examples:Examples: ArcSDE – Create GeoDatabase in RDBMSArcSDE – Create GeoDatabase in RDBMS SPIT – Connects QGIS to PostgreSQL(PostGIS)SPIT – Connects QGIS to PostgreSQL(PostGIS)

Open (Source) GISOpen (Source) GIS► PostGIS, MySQL – Open Source DBMS – implement OGC standardsPostGIS, MySQL – Open Source DBMS – implement OGC standards

► OGC – Consortium of 482 Companies/Orgs – Define OGC – Consortium of 482 Companies/Orgs – Define OpenOpen Standards Standards

► OSGeo Found’n – Support develop’t of op’n source Geospatial OSGeo Found’n – Support develop’t of op’n source Geospatial softwaresoftware

► GRASS – Orig. US Army, now project of OS GeoGRASS – Orig. US Army, now project of OS Geo

► GDAL (Conversion), GEOS (Geometry), rasdaman (rasters)GDAL (Conversion), GEOS (Geometry), rasdaman (rasters)

► Quantum GIS (QGIS) – Another OSGeo ProjectQuantum GIS (QGIS) – Another OSGeo Project MapServer export, OpenStreetMap editor, Run GRASS datasets/tools MapServer export, OpenStreetMap editor, Run GRASS datasets/tools

withinwithin

► MapTiler (uses GDAL2Tiles) – Create OpenLayers/Google Maps MapTiler (uses GDAL2Tiles) – Create OpenLayers/Google Maps TilesetsTilesets

Web GISWeb GIS

► Tools – MapTiler, OpenLayers, MapServer, GeoServerTools – MapTiler, OpenLayers, MapServer, GeoServer

► Developers’ Platforms – JavaScript, AJAX, SVG, JavaDevelopers’ Platforms – JavaScript, AJAX, SVG, Java

► Google Earth/Maps, Virtual Earth, StreetmapGoogle Earth/Maps, Virtual Earth, Streetmap

► OpenStreetMap, (interesting to compare against OS!)OpenStreetMap, (interesting to compare against OS!)

► OS Get-a-map; OS OpenData (Apr 2010); OpenData OS Get-a-map; OS OpenData (Apr 2010); OpenData APIAPI

► Simple User Requirements? – ArcGIS OnlineSimple User Requirements? – ArcGIS Online

The Future 1 – PDAs, AR, The Future 1 – PDAs, AR, AppsApps

►Move from mainframe to desktop to Move from mainframe to desktop to distributed desktops, and distributed distributed desktops, and distributed serversservers

►PDAs and Phones – ArcPad, GPS MapsPDAs and Phones – ArcPad, GPS Maps

►Augmented Reality – Scan horizon, UtilitiesAugmented Reality – Scan horizon, Utilities

►AIS – Ships, Planes – Track in Google EarthAIS – Ships, Planes – Track in Google Earth

The Future 2 – LBS, Sensors, The Future 2 – LBS, Sensors, Clouds Clouds

► Location Based Services, Sensor NetworksLocation Based Services, Sensor Networks

► Cloud Computing, Ubiquitous ComputingCloud Computing, Ubiquitous Computing

► Tracking People? Civil Liberties / Freedoms?Tracking People? Civil Liberties / Freedoms?

► Social inequality – advantaged vs disadvantaged?Social inequality – advantaged vs disadvantaged?

► Free/VGI => more folk making bad maps - Free/VGI => more folk making bad maps - noisenoise??

► But a more skilled-up knowledgeable population…?But a more skilled-up knowledgeable population…?

Thank You!Thank You!

Questions?Questions?