Summer Exam Flashcards
Geomatics definition
Modern scientific term referring to inter grated approach of measurement, analysis, management and storage and display of descriptions and location of earth based data- School of geomatics university of NSW.
Essentially application of GPS GIS and remote sensing
Remote sensing definition
Barrel and Curtis 1999- sciences of observation from a distance- VAGUE
Lillesland et al 2004- science and art of obtaining info about an object/ area through analysis of data acquired by device not in (physical) contact with object/ area under investigation.
GPS definition
System of orbiting satellites used for navigation purposes and capable of giving highly accurate geographic co-ordinates using hand held receivers- Heywood et al 2002
GIS definition
ESRI- an organised collection of computer hardware, software data and personally designed to efficiency store, update, manipulate and display all forms of geographically referenced information
āSā in GIS
System- means people, software, hardware and data.
Science- Issues surrounding use of GIS- data quality and interpretation.
Studies- legal and ethical use of GI
Service- NEWEST IDEA- GI to use Internet
GIS output
Maps, tables, graphs, images, 3D fly throughs.
Map output is perhaps most important form of spatial data representation. - layout important as to convey all relevant information.
Tackling a problem using geomatics example
What is the question? Is it getting wetter in Cumbria
What data are required? Rainfall data in time and space
How do I get data? How can data manipulated to answer question? Rainfall data- reformat it, map it- colour code accordingly then use it to answer question.
GPs systems
See sheet 2 for diagram.
Set of satellites which allow receivers to determine exact location 24 hours a day.
2 current systems: NAVSTAR (USA) GLONASS (Russia), European and Chinese systems under construction.
NAVSTAR- most commonly used: consists of 24 satellites each has 12 hour orbit. Pass over control systems so their orbits can be monitored and positions precisely identified.
GPS data transmission
- Radio signals are transmitted to receiver from each satellite.
- Time delay measured between transmission and receiving signal.
- Time delay, distance between receiver and satellite calculated from this.
- Signal from 3 or more satellites, triangulation can be calculated 2D location of receiver.
- 4 or more satellites (often more found but some ignored) , altitude of receiver can be calculated. Further apart satellites= more accurate position.
Dilution of precision of GPS
No GPS is 100% accurate, generally +/- 25m for 95% the time. Accuracy of position is measure by DOP- quantify and combine each source of error to calculate DOP.
Errors in GPS data
Ionosphere and troposphere delays:
Satellite signals slow as it passes through atmosphere. GPS system uses built in model that calculates an average amount of delay to partially correct for type of error.
Signal multipath:
GPS signal reflected off objects such as tall buildings increases travel of signal, causes errors.
Receiver clock errors:
Not as accurate as atomic clocks onboard GPS satellites may have slight timing errors.
Orbital errors:
Inaccuracy in satellites reported location
Intentional degradation of satellite signal:
Selective availability is intentional degradation of signal once I posed by US department of defence. Accuracy was decreased to 100m 95% of the time.
Differential GPS
Can be used to improve accuracy.
Techniques use 2 receivers:
Static base unit of precisely known location and roving receiver.
If both receivers are identically set up and used in close proximity, then positional error will be same for both units. In UK base states for differentiational correction are located in light houses.
GPS receiver
Shows relation to the WG884 data.
Most receivers convert this to any spatial reference system: British national grid.
Point data in GIS - explanation
Almost any feature can be represented as a point.
Extra non spatial information can be attached to that point and stored in a database.
Attributes give additional information about character of entities- good attribute makes a clear GIS output.
Applications of GIS point data
Surveying and field data collection.
Numerous applications in field- location of post offices etc
All geographers projects will benefit from mapping of features- a geographers shorthand.
Databases basic explanation
Collection of information that is organised that can be managed accessed and updated.
Developed by IBM in 50s in response to needs of military and business of government.
DBMS- database management systems- GIS is a competent of this
Software application that interacts with user, other applications and database itself to capture and analyse data eg Microsoft Acces.
Database concepts
User of application does not need to know how data is stored just how to access it- data transparency.
Changes can be made to database without effecting other components of system.
Database advantages
Reduction in data transparency, reduced software development costs, multiple external views the database depending on application.
Analysis only as good as attribute data.
Non data components of database
Operations: user functions for data manipulation eg sorting/ deleting records.
Data definition language: describe constants of data base- field attribute names.
Query language: standard language used to edit, analyse- input and output data.
File studies: several common data models- define how data is stored in system and they are connected to each other.
Data models- flat and relational files
Flat file- data on a single table- useful but simplistic cannot deal with relationship between objects- addressed using heirachical , network or relationship data.
Relational file- most popular data model and widely used in GIS. Each table has a key, common interger such as an ID appears in every ta le Iām database- several tables are easily linked.
Flat versus relation files
Flat models can be seen as cumbersome and lead to data repetition.
Relational models are: efficient in terms of storage and access, easy to update, allow different users to access selected information, more consistent with a greater integrity of data.
In flat file, both attribute data and spatial data in same file.
Relational model is more common where spatial data is stored as part of tables.
Vector topology definition
Heywood et al 2002- geomatics relationship between objects located in space.
Topology consist of 3 elements- Burough 1986- adjacency, containment and connectivity. Challenge is to mantain topology using minimum amount of data
Vector point data
Simplist type of spatial object.
Choice of entities to represent as points depends on scale of map- from individual buildings to cities.
If height is known, this can be added as an attribute effectively describing the point as an co ordinate- effectively uses database flat file.
Vector polylines data
Series of connected points.
Polylines used in GIS infrastructure (roads, railways, gas, electricity)
Polylines constructed of georefrenced nodes and links.
Link attribute- streets and pipe width
Node attribute- traffic lights etc- each node can be assigned a z value to denote height and make model 3D.
Vector polylines networks
Networks are an extension of polylines.
Consist interconnected polylines that allow for flow of objects: hydrology, transpiration and information.
Requires good attribute data eg. Route planing needs in for re one way streets.
Using vector data
Can pretty much model any geographical features, series of features built up as layers to produce and interactive map. Each layer contains a difficult spatial entity.
Advantages if vector data
Well suited to map output.
Resolution independent of detail.
Efficient encoding for topology.
Issues with vector data
How to represent change over time?
What features are portrayed depends on concoetualisation, scale and application.
Raster data description
Used to represent continuous data- divided into rows and columns grid- individual elements called cells.
Digital photos common.
Raster data can hold any attribute value eg elevation, vegetation.
Positives of raster data
Simple data structure.
High spatial variability is efficiency represented.
Raster is good for storing image data.
Negatives of raster data
Storage size is a big issue- large array of values.
Topology hard to represent.
Cell size can be issue- maps may appear blocky
Digital terrain models- DTM- raster data
Surfaces can be modelled using vector- more common to use raster.
Digital representation of a topographic surface- grid of spot heights.
Data from: satellite images, existing maps and ground surveys.
DTM created using algorithms which interpolate points or contour lines.
DTM issues
Assumes height is costant across cell.
High resolutions are required for accuracy, this is inappropriate for others leading to too much data,
Remote sensing- radiation in atmosphere
Radiation sensed by satellite either reflected from surface or emitted from surface, all bodies greater than -273 degrees centigrade emit radiation.
Electromagnetic spectrum
Remotely collected data can take many forms, but term remit sensing specifically refers to sensing of electromagnetic energy.
Thermal energy and microwaves emitted.
Spectral reflectance in remote sensing
Each surface has own signature of reflectance.
For every wavelength of electromagnetic spectrum.
Built up to create spectral reference curve.
Sensors and satellites- remote sensing
Satellites have multitude of sensors on board, each targeting different section of electromagnetic spectrum.
Satellite imaging
Sensor will produce raster image,
Build up raster image via a swath width.
With width mirror reflects on image of ground, pixel by pixel- pixel size and therefore resolution varies.
Image is a mixture of digital numbers, each number- degree of reflectance
Combine bands
Eg NDVI- normalised difference vegetation index.
Healthy veg absorbs most visible light (VIS) and reflects near infra red (NIR)
Remote sensing and DEMS
DEMS create by using remote sensing techniques.
Very high resolution and accurate to +/- 15 cm.
Positive of remote sensing
Provides a unique opportunity to observe large areas in detail.
Repeatable and temporal resolution.
Wide ra he of applications.
Cheap for end user.
Disadvantages of remote sensing
Deployment of satellites very expensive.
Doesnāt produce measurements just estimates from observations.
Sources of data- metadata
Data about data.
Essential to allow for efficient search and discover difference between users- libraries use metadata for books.
Langley et al 2001- for spatial data, the metadata must source many different purposes, from descriptions of contents for handling.
ArcGIS online
Hosted by ESRI and new content is added and updated continually. GIS users can access freely through Arc GIS online as map service. Internet often has compromise: lower resolution data only part processe.
Raw satellite data
Vector data may often not be available for many regions of globe.
Often you will have to resort to satellite data, which you can then process.
Comprehensive amount of data now freely available.
Satellite data issues
Data is raw- also very compressed and need specialising.
Pictures of whole globe and so projection will not immediately be suitable for UK.
Map projections
Despite 3D GIS projections still needed for imagery.
World needs to be treated as flat- image distorted as it is squashed. Method known as projection.
Projections
Cylindrical projection- continuous photo of earth, countries near equator in true positions, view of poles distorted, area mostly preserved.
Azimuthal projection- only a part of earths surface is visible, view half of globe, distortion at 4 edges but distance mostly preserved.
Conic- area distorted, distance distorted to bottom of image, scale preserved.
Sinusodial- preserves area, used frequently as standard satellite data, not easy to correct for use in UK.
Spatial referencing
Used to locate a feature on 2D surface.
Methods george aground co ordinate systems (latitude and longitude), rectangular co ordinates.
Rectangular co ordinates- projections
Map made by projecting lines of lat and long onto flat map.
Error of distortion smaller over lower area, globe split into rectangles grids. Eg British national grid grid from Scilly to Shetland.
Ordnance survey- free to use data, funded by tax payer.
Digital- co ordination of online mapping sites such as street map or multi map.
Photogrammetry- projections
Areal photos primitive method of remote sensing.
Wide availability in time and space and wide area view. Spatial resolution. However scale varies across image and no spatial referencing.
Stereo imaging
Distortion in vertical images allows to be seen in 3D- stereo pair images offset from each other- results in a parallax.
Manual digitising have 2 modes
Point mode- lines are generalised as series of points defined by user.
Stream mode- lines generalise by points at set individuals.
Error- generalisation, human and incorrect registration.
Automatic digitising
Quicker than manual.
Uses a scanner- located into GIS for digitising.
Error- optical distortion from scanner, scanning unwanted info, resolution is only what scanner sets it to.
Georegistering- editing
Process of adjusting a map/ photo / image to geographic location known āgoodā image.
Data needs a common projection co ordinate system and scale.
Remote sensing errors in editing
Random errors occur in satellite data.
Complexity if the science leaves room for error.
Radio metric and geometric errors most common.
Remote sensing errors: radio metric and geometric error
R- problems measuring radiance eg reflection and atmospheric conditions.
G- rotation of earth during images causes distortion. Why satellite images never square as lines cause geometric error which must be pulled back, curvature of earth also distorts image.
Rubber sheeting
Used to adjust features in target in non uniform manner.
Involves stretches target as if it was a rubber sheet.
More ground control points = more accurate sketch.
Automatic vectorisation
Best used on single targets,
Converts raster data to vector data- little say in what is digitised.
GIS data quality terms
Error- physical difference between GIS and real world
Accuracy- estimated value from true value
Precision- recorded level of detail. High precision not equal to high accuracy.
Bias- system to variation of data from reality
Resolution- raster datasets
Completeness- data set is complete spatially and temporally.
Compatibility- refers to consistent project, scale and co ordinate system.
Applicability- can dataset be used to solve problem?
Common spatial data errors
Missing entities and missing points, lines or boundaries.
Duplicate entities/ digitised twice.
Artefacts of digitising- wrongly placed nodes
Raster analysis of a surface
Heywood et al 2002 ācontinuous variation in space of 3rd dimensionā
Surface analysis creates new output from one or more input rasters.
Raster analysis: contrast stretching
Image enhancement- often done before more complicated analyses.
Enables more detail to be seen.
DN of pixels plotted as a histogram which is then stretched.
Raster analysis: reclassification
Simple raster analysis, queuing vector data, used frequently with satellite imaging to identify geographical areas simple rules such as:
Old cell value
Raster analysis: neighbourhood functions
Concerned with processing adjacent values of cells.
2 types: proximity and filtering.
Analysis produces new raster layer where output raster consists of cells contains distance data.
Filtering changes value of target cell based on values of neighbouring cells. Filter can be varied in several ways: size, shape, main function. Can be used for image sharpening and edge detection.
Neighbourhood slope: raster analysis
Steepness of terrain, lower value is a flatter terrain.
Aspect us direction in which terrain faces.
Viewsheds: raster analysis
Also known as viability analysis.
Used for location of wind farms
Visualised with a fly through. Calculated using Ray tracing.
Sophistication can be added by including obstructions such as trees in analysis.
Multiple rasters: raster analysis
Analysis often requires processing of several rasters.
Can cause issue of rasters not all same resolutions. Best resolution obtained from analysis will be same as other rasters.
Nature of satellites data often means dealing with more than one raster.
Raster analysis: colour composites
Create a colour composite you need to combine 3 bands.
Any colour made from red, green or blue.
Satellite draping: raster analysis
Given to procedure when surface us overlain by another layer.
Satellite imagery can be draped over DEM add more geographical detail.
Terrain values can be incorporated into other raster as additional attribute data.
Raster data analysis: image classification
Certain combinations of rasters can be very useful to determine cover types.
Unsupervised classification, computer picks centres of clusters and divides pixels accordingly.
Supervised classification: need to train computer, use fieldwork to identify pixels to teach computer. Computer looks for pixels with similar spectral properties.
Raster data integration
Integration easier with raster than vector data.
Berry 1993 overlaying and combining data in different raster layers is called map algobia - preferred on individual cells on 2 or more layers to produce new layer.
Mathematic modelling- raster analysis
Multiple regression can be applied to series of rasters.
Data can also be extracted from GIS in a table so equations can be derived in excel and SPSS- predictive statistics.
Vector analysis: queries
Common database function for data retrieval.
A spatial queries ask about attribute data.
Spatial use GIS functions.
Queries can also be combined using Booleon operators- using wizards to get through process.
Vector analysis: buffering
Very common function- width user defined
Draw zone of interest around vector entity.
Output in form of polygon layer.
Vector analysis- data integration
Sometimes known as geoprocessing
Uses layer concept to overlay different datasets and integrate into a single map.
Instead of just visualising results of query a new layer is produced.
Vector analysis: dissolve and merge
Dissolve- Used to aggregate features with same selected attribute.
Merge-allows adjacent layers to merge. Speeds up analysis as queries and other integrations are only one layer.
Vector analysis: spatial joining
Spatial analysis in which attributes of features in 2 different layers are joined together based in relative location of feature.
Tolerance can be set to limit extent of joining or join can be made- distance of join recorded as attribute.
Vector analysis: vector data integration
Often necessary to combine many different processes to achieve acquired. Topology of datasets have to be seen to avoid dudd results
Vector to raster and raster to vector data
Often quicker to perform analysis with raster rather than vector data- analysis also easier if all layers vector or all raster.
Raster into vector- vectorisation
Vector into raster- rasterisation
Vector analysis: spatial interpolation
Useful feature of GIS
Concoetually easy- theory complex however.
Waters 1989- spatial interpolation is procedure of estimating the values of properties at unsampled sites within an area covered by existing observations.
Vector analysis: spatial interpolation
Elevation models provide excellent example- contour lines produced by eyeballing spot heights- time consuming and assumes slope constant between points.
Density function: vector analysis
Map concentration of polyline- values spread across area and density value is calculated for each cell in the raster.
Geomatics definition
Modern scientific term referring to inter grated approach of measurement, analysis, management and storage and display of descriptions and location of earth based data- School of geomatics university of NSW.
Essentially application of GPS GIS and remote sensing
Remote sensing definition
Barrel and Curtis 1999- sciences of observation from a distance- VAGUE
Lillesland et al 2004- science and art of obtaining info about an object/ area through analysis of data acquired by device not in (physical) contact with object/ area under investigation.
GPS definition
System of orbiting satellites used for navigation purposes and capable of giving highly accurate geographic co-ordinates using hand held receivers- Heywood et al 2002
GIS definition
ESRI- an organised collection of computer hardware, software data and personally designed to efficiency store, update, manipulate and display all forms of geographically referenced information
āSā in GIS
System- means people, software, hardware and data.
Science- Issues surrounding use of GIS- data quality and interpretation.
Studies- legal and ethical use of GI
Service- NEWEST IDEA- GI to use Internet
GIS output
Maps, tables, graphs, images, 3D fly throughs.
Map output is perhaps most important form of spatial data representation. - layout important as to convey all relevant information.
Tackling a problem using geomatics example
What is the question? Is it getting wetter in Cumbria
What data are required? Rainfall data in time and space
How do I get data? How can data manipulated to answer question? Rainfall data- reformat it, map it- colour code accordingly then use it to answer question.
GPs systems
See sheet 2 for diagram.
Set of satellites which allow receivers to determine exact location 24 hours a day.
2 current systems: NAVSTAR (USA) GLONASS (Russia), European and Chinese systems under construction.
NAVSTAR- most commonly used: consists of 24 satellites each has 12 hour orbit. Pass over control systems so their orbits can be monitored and positions precisely identified.
GPS data transmission
- Radio signals are transmitted to receiver from each satellite.
- Time delay measured between transmission and receiving signal.
- Time delay, distance between receiver and satellite calculated from this.
- Signal from 3 or more satellites, triangulation can be calculated 2D location of receiver.
- 4 or more satellites (often more found but some ignored) , altitude of receiver can be calculated. Further apart satellites= more accurate position.
Dilution of precision of GPS
No GPS is 100% accurate, generally +/- 25m for 95% the time. Accuracy of position is measure by DOP- quantify and combine each source of error to calculate DOP.
Errors in GPS data
Ionosphere and troposphere delays:
Satellite signals slow as it passes through atmosphere. GPS system uses built in model that calculates an average amount of delay to partially correct for type of error.
Signal multipath:
GPS signal reflected off objects such as tall buildings increases travel of signal, causes errors.
Receiver clock errors:
Not as accurate as atomic clocks onboard GPS satellites may have slight timing errors.
Orbital errors:
Inaccuracy in satellites reported location
Intentional degradation of satellite signal:
Selective availability is intentional degradation of signal once I posed by US department of defence. Accuracy was decreased to 100m 95% of the time.
Differential GPS
Can be used to improve accuracy.
Techniques use 2 receivers:
Static base unit of precisely known location and roving receiver.
If both receivers are identically set up and used in close proximity, then positional error will be same for both units. In UK base states for differentiational correction are located in light houses.
GPS receiver
Shows relation to the WG884 data.
Most receivers convert this to any spatial reference system: British national grid.
Point data in GIS - explanation
Almost any feature can be represented as a point.
Extra non spatial information can be attached to that point and stored in a database.
Attributes give additional information about character of entities- good attribute makes a clear GIS output.
Applications of GIS point data
Surveying and field data collection.
Numerous applications in field- location of post offices etc
All geographers projects will benefit from mapping of features- a geographers shorthand.
Databases basic explanation
Collection of information that is organised that can be managed accessed and updated.
Developed by IBM in 50s in response to needs of military and business of government.
DBMS- database management systems- GIS is a competent of this
Software application that interacts with user, other applications and database itself to capture and analyse data eg Microsoft Acces.
Database concepts
User of application does not need to know how data is stored just how to access it- data transparency.
Changes can be made to database without effecting other components of system.
Database advantages
Reduction in data transparency, reduced software development costs, multiple external views the database depending on application.
Analysis only as good as attribute data.
Data types in database.
Numerical: intergers, scales, domains can set range of possible values.
Non numerical: Characters dates and logic (true/ false)
See table on sheet 3
Non data components of database
Operations: user functions for data manipulation eg sorting/ deleting records.
Data definition language: describe constants of data base- field attribute names.
Query language: standard language used to edit, analyse- input and output data.
File studies: several common data models- define how data is stored in system and they are connected to each other.
Data models- flat and relational files
Flat file- data on a single table- useful but simplistic cannot deal with relationship between objects- addressed using heirachical , network or relationship data.
Relational file- most popular data model and widely used in GIS. Each table has a key, common interger such as an ID appears in every ta le Iām database- several tables are easily linked.
Flat versus relation files
Flat models can be seen as cumbersome and lead to data repetition.
Relational models are: efficient in terms of storage and access, easy to update, allow different users to access selected information, more consistent with a greater integrity of data.
In flat file, both attribute data and spatial data in same file.
Relational model is more common where spatial data is stored as part of tables.
Spatial data over view
Data models: topology
2 distinct data models used- vector and raster
Vector model: points, polygons and polylines and surfaces.
Point data- x,y- building block of vector data.
GIS software designed to use spatial data.
Vector topology definition
Heywood et al 2002- geomatics relationship between objects located in space.
Topology consist of 3 elements- Burough 1986- adjacency, containment and connectivity. Challenge is to mantain topology using minimum amount of data
Vector point data
Simplist type of spatial object.
Choice of entities to represent as points depends on scale of map- from individual buildings to cities.
If height is known, this can be added as an attribute effectively describing the point as an co ordinate- effectively uses database flat file.
Vector polylines data
Series of connected points.
Polylines used in GIS infrastructure (roads, railways, gas, electricity)
Polylines constructed of georefrenced nodes and links.
Link attribute- streets and pipe width
Node attribute- traffic lights etc- each node can be assigned a z value to denote height and make model 3D.
Vector polylines networks
Networks are an extension of polylines.
Consist interconnected polylines that allow for flow of objects: hydrology, transpiration and information.
Requires good attribute data eg. Route planing needs in for re one way streets.
Vector polygon data
Essentially area data. Formed from a closed chain of points/ nodes. Island- a woodland Adjacent- counties Nested- contour lines
3 types of attribute node, link and area.
Some scientists consider 4th entity type. Continuous surfaces include elevation, rainfall, temp and pop.
Using vector data
Can pretty much model any geographical features, series of features built up as layers to produce and interactive map. Each layer contains a difficult spatial entity.
Advantages if vector data
Well suited to map output.
Resolution independent of detail.
Efficient encoding for topology.
Issues with vector data
How to represent change over time?
What features are portrayed depends on concoetualisation, scale and application.
Raster data description
Used to represent continuous data- divided into rows and columns grid- individual elements called cells.
Digital photos common.
Raster data can hold any attribute value eg elevation, vegetation.
Positives of raster data
Simple data structure.
High spatial variability is efficiency represented.
Raster is good for storing image data.
Negatives of raster data
Storage size is a big issue- large array of values.
Topology hard to represent.
Cell size can be issue- maps may appear blocky
Digital terrain models- DTM- raster data
Surfaces can be modelled using vector- more common to use raster.
Digital representation of a topographic surface- grid of spot heights.
Data from: satellite images, existing maps and ground surveys.
DTM created using algorithms which interpolate points or contour lines.
DTM issues
Assumes height is costant across cell.
High resolutions are required for accuracy, this is inappropriate for others leading to too much data,
DTM derivatives- many ego applications
Climatological- slope, aspect and hillshade
Planning- wind farms and development.
Remote sensing- radiation in atmosphere
Radiation sensed by satellite either reflected from surface or emitted from surface, all bodies greater than -273 degrees centigrade emit radiation.
Electromagnetic spectrum
Remotely collected data can take many forms, but term remit sensing specifically refers to sensing of electromagnetic energy.
Thermal energy and microwaves emitted.
Spectral reflectance in remote sensing
Each surface has own signature of reflectance.
For every wavelength of electromagnetic spectrum.
Built up to create spectral reference curve.
Sensors and satellites- remote sensing
Satellites have multitude of sensors on board, each targeting different section of electromagnetic spectrum.
Satellite imaging
Sensor will produce raster image,
Build up raster image via a swath width.
With width mirror reflects on image of ground, pixel by pixel- pixel size and therefore resolution varies.
Image is a mixture of digital numbers, each number- degree of reflectance
Combine bands
Eg NDVI- normalised difference vegetation index.
Healthy veg absorbs most visible light (VIS) and reflects near infra red (NIR)
Remote sensing and DEMS
DEMS create by using remote sensing techniques.
Very high resolution and accurate to +/- 15 cm.
Positive of remote sensing
Provides a unique opportunity to observe large areas in detail.
Repeatable and temporal resolution.
Wide ra he of applications.
Cheap for end user.
Disadvantages of remote sensing
Deployment of satellites very expensive.
Doesnāt produce measurements just estimates from observations.
Sources of data- metadata
Data about data.
Essential to allow for efficient search and discover difference between users- libraries use metadata for books.
Langley et al 2001- for spatial data, the metadata must source many different purposes, from descriptions of contents for handling.
Spatial metadata standards
Federal geographic data committee attempted to standardise made it too complicated so it was never used. Arc Catalog general geographic data organiser.
ArcGIS online
Hosted by ESRI and new content is added and updated continually. GIS users can access freely through Arc GIS online as map service. Internet often has compromise: lower resolution data only part processe.
Raw satellite data
Vector data may often not be available for many regions of globe.
Often you will have to resort to satellite data, which you can then process.
Comprehensive amount of data now freely available.
Satellite data issues
Data is raw- also very compressed and need specialising.
Pictures of whole globe and so projection will not immediately be suitable for UK.
Map projections
Despite 3D GIS projections still needed for imagery.
World needs to be treated as flat- image distorted as it is squashed. Method known as projection.
Projections
Cylindrical projection- continuous photo of earth, countries near equator in true positions, view of poles distorted, area mostly preserved.
Azimuthal projection- only a part of earths surface is visible, view half of globe, distortion at 4 edges but distance mostly preserved.
Conic- area distorted, distance distorted to bottom of image, scale preserved.
Sinusodial- preserves area, used frequently as standard satellite data, not easy to correct for use in UK.
Spatial referencing
Used to locate a feature on 2D surface.
Methods george aground co ordinate systems (latitude and longitude), rectangular co ordinates.
Rectangular co ordinates- projections
Map made by projecting lines of lat and long onto flat map.
Error of distortion smaller over lower area, globe split into rectangles grids. Eg British national grid grid from Scilly to Shetland.
Ordnance survey- free to use data, funded by tax payer.
Digital- co ordination of online mapping sites such as street map or multi map.
Photogrammetry- projections
Areal photos primitive method of remote sensing.
Wide availability in time and space and wide area view. Spatial resolution. However scale varies across image and no spatial referencing.
Stereo imaging
Distortion in vertical images allows to be seen in 3D- stereo pair images offset from each other- results in a parallax.
Manual digitising have 2 modes
Point mode- lines are generalised as series of points defined by user.
Stream mode- lines generalise by points at set individuals.
Error- generalisation, human and incorrect registration.
Automatic digitising
Quicker than manual.
Uses a scanner- located into GIS for digitising.
Error- optical distortion from scanner, scanning unwanted info, resolution is only what scanner sets it to.
Georegistering- editing
Process of adjusting a map/ photo / image to geographic location known āgoodā image.
Data needs a common projection co ordinate system and scale.
Remote sensing errors in editing
Random errors occur in satellite data.
Complexity if the science leaves room for error.
Radio metric and geometric errors most common.
Remote sensing errors: radio metric and geometric error
R- problems measuring radiance eg reflection and atmospheric conditions.
G- rotation of earth during images causes distortion. Why satellite images never square as lines cause geometric error which must be pulled back, curvature of earth also distorts image.
Rubber sheeting
Used to adjust features in target in non uniform manner.
Involves stretches target as if it was a rubber sheet.
More ground control points = more accurate sketch.
Automatic vectorisation
Best used on single targets,
Converts raster data to vector data- little say in what is digitised.
GIS data quality terms
Error- physical difference between GIS and real world
Accuracy- estimated value from true value
Precision- recorded level of detail. High precision not equal to high accuracy.
Bias- system to variation of data from reality
Resolution- raster datasets
Completeness- data set is complete spatially and temporally.
Compatibility- refers to consistent project, scale and co ordinate system.
Applicability- can dataset be used to solve problem?
Common spatial data errors
Missing entities and missing points, lines or boundaries.
Duplicate entities/ digitised twice.
Artefacts of digitising- wrongly placed nodes
Raster analysis of a surface
Heywood et al 2002 ācontinuous variation in space of 3rd dimensionā
Surface analysis creates new output from one or more input rasters.
Raster analysis: contrast stretching
Image enhancement- often done before more complicated analyses.
Enables more detail to be seen.
DN of pixels plotted as a histogram which is then stretched.
Raster analysis: reclassification
Simple raster analysis, queuing vector data, used frequently with satellite imaging to identify geographical areas simple rules such as:
Old cell value
Neighbourhood slope: raster analysis
Steepness of terrain, lower value is a flatter terrain.
Aspect us direction in which terrain faces.
Viewsheds: raster analysis
Also known as viability analysis.
Used for location of wind farms
Visualised with a fly through. Calculated using Ray tracing.
Sophistication can be added by including obstructions such as trees in analysis.
Multiple rasters: raster analysis
Analysis often requires processing of several rasters.
Can cause issue of rasters not all same resolutions. Best resolution obtained from analysis will be same as other rasters.
Nature of satellites data often means dealing with more than one raster.
Raster analysis: colour composites
Create a colour composite you need to combine 3 bands.
Any colour made from red, green or blue.
Satellite draping: raster analysis
Given to procedure when surface us overlain by another layer.
Satellite imagery can be draped over DEM add more geographical detail.
Terrain values can be incorporated into other raster as additional attribute data.
Raster data analysis: image classification
Certain combinations of rasters can be very useful to determine cover types.
Unsupervised classification, computer picks centres of clusters and divides pixels accordingly.
Supervised classification: need to train computer, use fieldwork to identify pixels to teach computer. Computer looks for pixels with similar spectral properties.
Raster data integration
Integration easier with raster than vector data.
Berry 1993 overlaying and combining data in different raster layers is called map algobia - preferred on individual cells on 2 or more layers to produce new layer.
Vector analysis: queries
Common database function for data retrieval.
A spatial queries ask about attribute data.
Spatial use GIS functions.
Queries can also be combined using Booleon operators- using wizards to get through process.
Vector analysis: buffering
Very common function- width user defined
Draw zone of interest around vector entity.
Output in form of polygon layer.
Vector analysis- data integration
Sometimes known as geoprocessing
Uses layer concept to overlay different datasets and integrate into a single map.
Instead of just visualising results of query a new layer is produced.
Vector analysis: dissolve and merge
Dissolve- Used to aggregate features with same selected attribute.
Merge-allows adjacent layers to merge. Speeds up analysis as queries and other integrations are only one layer.
Vector analysis: spatial joining
Spatial analysis in which attributes of features in 2 different layers are joined together based in relative location of feature.
Tolerance can be set to limit extent of joining or join can be made- distance of join recorded as attribute.
Vector analysis: vector data integration
Often necessary to combine many different processes to achieve acquired. Topology of datasets have to be seen to avoid dudd results
Vector to raster and raster to vector data
Often quicker to perform analysis with raster rather than vector data- analysis also easier if all layers vector or all raster.
Raster into vector- vectorisation
Vector into raster- rasterisation
Vector analysis: spatial interpolation
Useful feature of GIS
Concoetually easy- theory complex however.
Waters 1989- spatial interpolation is procedure of estimating the values of properties at unsampled sites within an area covered by existing observations.
Vector analysis: spatial interpolation
Elevation models provide excellent example- contour lines produced by eyeballing spot heights- time consuming and assumes slope constant between points.
Density function: vector analysis
Map concentration of polyline- values spread across area and density value is calculated for each cell in the raster.