Lecture 5 Flashcards
GIS data sets often contain 3 key components:
- A spatial reference system
- Attribute data linked to the spatial features
- Metadata
- if there is no Metadata either a good sign of not great data or you have to ask why they didn’t include it
Accuracy Vs. Precision
Accuracy:the extent to which both attribute and positional data correspond to their real-world counterparts
eg, your road network data set has an accuracy of ±0.25 m – your data set is within 0.25 m of matching the real world road network
Precision: the exactness or repeatability of the measurements
-often described as the number of decimal places you can report your measurements to (compare 0.5 to 0.48739
T OR F
high precision DOES NOT EQUAL high accuracy and the same reversed
T
▪ consider a car GPS system that boasts precise locating to less than 1 metre, but doesn’t have an accurate map of your location
Types of errors when collecting measurments: 6
▪ attribute error (including logical consistency and completeness)
▪ positional error (x, y, and z)
▪ topological geometric error
▪ temporal error
▪interpretation error due to ecological fallacy
▪ error due to the modifiable areal unit problem
2 common methods used to determine attribute accuracy:
1.random spot-checking
▪ in very large databases, checking every attribute is impractical; randomly choose 2% or fewer records and manually check and assess errors
2.spatial sampling
▪ use basic sampling techniques to choose how many records to check from each defined subset
An error matrix quickly displays the error
▪ for categorical data (nominal or ordinal) assess errors in attribute data
for finding error in interval and ratio data
the root-mean-square-error(RMSE) is commonly used
▪ RMSE compares the value of an attribute at a given location and compares it to the real world value at the same location
▪ consider a feature map showing leaf area idex (LAI)
Errors:
- attribute logical consistency
- attribute and spatial completeness
1.
▪ in many cases, 2 or more people are involved in the collection of attribute data
▪ in this case, it is important that those collectors follow the same rules and logic for data collection to ensure that the whole dataset is consistent
▪ systematic errors occur when one set of data is consistently off of another set, usually because of some sort of user induced error
2.
▪ completeness is the degree to which the data exhaust the universe of all possible items
▪ a spatially complete data set covers all of the area of interest with the same level of detail
▪ hard to do since some locations are easier to access and collect data from
▪ a thematically (attribute) complete data set ensures that the data contain all the thematic information necessary for a given project and that it was collected with the same level of precision
▪ incomplete field, partial records, or missing records can wreak havoc with GIS analysis, especially if calculated statistics are involved
topological errors
topological errors are usually created inside the GIS by the user as the geodatabase is being built
▪ a vector geodatabase is composed of lines and polygons – lines may not intersect or overshoot, and polygons may not close
▪ a fuzzy tolerance may be used to correct topological errors, but selecting the specific fuzzy tolerance is tricky, since too large a tolerance can join distinct but close features
Shadow Map
A shadow map(slide 24)
-shows a map of error
-defines the confidence that you have in certain portions of a map
. using numbers or statistics to describe error is fine, but often visualizing those errors is a more effective way of describing them
ecological fallacy
.the belief that all observations within an area will exhibit the same or similar values for a particular characteristic
-the characteristics of the group are the same for each individual in the group
Modifiable Unit Area Problem
spatial data may be reported, mapped, and analyzed using enumeration units or districts of various sizes, including countries, regions, provinces, cities, census tracts, neighbourhoods, etc.
▪ when smaller areal units are combined into fewer but larger units,The variation present in the smaller units may decrease
. in general, the correlation between units increases as the size of the units increase, and it is possible for a researcher to manipulate the results of an analysis through this modifiable unit area problem
GIS is simply a tool to help us ______, _____, _____, ______, and ________ various types of spatial data
store, organize, manipulate, analyze and present
GIS DESIGN: 3 ESSENTIAL STEPS
- Identify Problem: every GIS application is designed to address a specific challenge or problem – for a GIS application to be successful, you must make sure that you fully understand the nature of the problem you are seeking to address
- Design and choose a data model: successful GIS applications model the part of the real world which is of interest to the problem under investigation
- Analysis Design: all GIS application are designed to allow users to explore a problem – the key to a successful analysis of a problem is the design of an appropriate framework for the analysis all GIS application are designed to allow users to explore a problem
GIS DESIGN-Problem identification
.either structured or unstructured problems
.develop root definition after using “rich picture”
.identify what resources are available
