GIS Ketepatan Data - Week 9

8/10/2019 GIS Ketepatan Data - Week 9

1/20

Data Accuracy

Definisi

Precision & Accuracy

Sources of Error

Error Propagation & Cascading


2/20

Definisi

Knowing the quality of the data is criticalto judging the applications for which theyare appropriate

Ironically, error arises from one ofgreatest strengths of GIS - the ability tocollate and cross-reference many types ofdata by location

Every time a new dataset is imported, theGIS also inherits its errors which reactwith existing errors in unpredictable ways


3/20

Precision & Accuracy


4/20

Accuracy

the degree to which information on a map or ina digital database matches true or acceptedvalues

pertains to the quality of data and the number

of errors contained in a dataset or mapincluding: horizontal and vertical accuracy with respect

to geographic position attribute, conceptual, and logical accuracy

the level of accuracy required for particular

applications varies greatly highly accurate data can be very difficult and

costly to produce and compile


5/20

Precision

refers to the level of measurement andexactnessof description in a GIS database

precise locational data may measure position

to a fraction of a unit precise attribute information may specify the

characteristics of features in great detail

precise data--no matter how carefullymeasured--may be inaccurate eg.

surveyors may make mistakes or data may beentered into the database incorrectly


6/20

the level of precision required for particularapplications varies greatly engineering projects eg. road and utility

construction, require very precise information- mm

demographic analyses of marketing orelectoral trends can often make do with less,say to the closest postcode or CD

highly precise data can also be very difficultand costly to collect

carefully surveyed locations needed by utilitycompanies to record the locations of pumps,wires, pipes and transformers cost $5-$20 perpoint to collect


7/20

High precision does not indicate highaccuracy nor does high accuracy implyhigh precision

Highly accurate and highly precise spatialinformation is not necessary for every GISapplication excessive accuracy and precision is not only

costly but can swamp your application inextraneous detail

Be aware also that GIS practitioners arenot always consistent in their use of theseterms


8/20

Additional terms

Two additional terms are used aswell:

Data Qualityrefers to the relativeaccuracy andprecision of a particularGIS database - often documented indata quality reports

Errorencompasses both theimprecision of data and its inaccuracies


9/20

Positional Accuracy

the expected deviance in the geographiclocation of a feature in the data set from itstrue location

tested by selecting a random sample of pointsin the data and comparing their coordinates totheir ground positions

U.S. Geological Survey Accuracy Standard:90% of all measurable points must be within

1/30th of an inch for maps at a scale of1:20,000 or larger, and 1/50th of an inch formaps at scales smaller than 1:20,000


10/20

Attribute Accuracy

the fidelity of the non-spatial elements of thedataset

inaccuracies may result from mistakes of manykinds

the non-spatial data itself can also vary greatlyin precision

example:

a precise description of a person living at aparticular address might include gender, age,

income, occupation, level of education, andmany other characteristics

an imprecise description might include justincome, or just gender


11/20

Logical Consistency

how well logical relations among dataelements are maintained

example: the edge of a property that

borders a lake should coincide with thelake boundary


12/20

Conceptual Accuracy

inaccuracies inherent in the conceptual designof the database

users may use inappropriate categories ormisclassify information

examples:

classifying cities by voting behavior wouldprobably be an ineffective way to studyfertility patterns

failing to classify power lines by voltage wouldlimit the effectiveness of a GIS designed tomanage an electric utilities infrastructure


13/20

Example:

A comparison of elevation

accuracies derived from DifferentialGPS and an airborne laserprofilometer
http://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htmhttp://watleo.uwaterloo.ca/~piwowar/Research/DEM_CRSS/index.htm


14/20

Sources of ErrorStage Sources of Error

Data Collection

age of data

incomplete data coverage

inappropriate map scales

use of "surrogate" data in place of non-existent primary data

errors in field data collection

errors in existing maps used as source data

errors in the analysis of remotely sensed imagery

Data Input inaccuracies in digitizing

inaccuracies inherent in the goegraphic features (e.g., "fuzzy" boundaries drawn as sharp lines)

Data Storageinappropriate data format

insufficient numerical precision

unsufficient spatial resolution

Data Manipulation

numerical computaton errors

inappropriate class intervalsboundary errors

error propogation as multiple overlays are combined

slivers created during polygon overlay

Data Outputscaling inaccuracies

output device inaccuracies

instability of the output medium

Use of Results

the information may be incorrectly understoodthe information may be inappropriately used


15/20

Error Propagation & Cascading

GIS usually involve operations onmanysets of dataInaccuracy, imprecision, and errormay be compounded in GIS thatemploy many data sources


16/20

Propagation

propagation occurs when one error leads toanother

example: if a map registration point has beenmis-digitised in one coverage and is then usedto register a second coverage, the secondcoverage will propagate the first mistake

in this way, a single error may lead to othersand spread until it corrupts data throughoutthe entire GIS project

to avoid this problem use the largest scale

map to register your points often propagation occurs in an additive

fashion, as when maps of different accuracyare collated


17/20


18/20

Cascading

o cascading occurs when errors areallowed to propagate unchecked fromlayer to layer repeatedly


19/20


20/20

the effects of cascading can be verydifficult to predict

can be additive or multiplicative

vary depending on how information iscombined

GIS Ketepatan Data - Week 9

Documents