SUKDANAN DIAGNOSTIC TEST IN PHYSICAL SCIENCE ANSWER KEYY.pdf
Tutorialinterpolationam en
1. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-1
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
TUTORIAL: INTERPOLATION OF
HYDROLOGICAL VARIABLES IN ARCGIS
SPATIAL ANALYST AND GEOSTATISTICAL
ANALYST
Josef Fürst, IWHW-BOKU
1 PROBLEM
Conductivities and Chloride content of soil water have been measured in several
observation points. These data shall be analysed and appropriately interpolated. We are
supposed to assess the influence of the selected interpolation method on the resulting
map.
2 TRAINING OBJECTIVES
• Display a map of our observed data
• As a first step, visualise areal characteristics by a THIESSEN map
• Deterministic interpolation methods: IDW and splines
• Geostatistical methods: Ordinary and Universal Kriging
• Comparison of methods
3 GIVEN DATA
Our data are recorded in a simple Dbase table sample1.dbf. Column ID is a unique point
identifier, columns X and Y are coordinates and Lf is conductivity, Cl is Chloride. A few
lines of this table are displayed below:
4 SOFTWARE REQUIREMENTS
ArcGIS 9.2 or later
2. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-2
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
Extensions: Spatial analyst, Geostatistical Analyst
5 TASKS
5.1 Display a map of the sample points
• Start ArcMap and load extensions listed above (if not already active)
• add sample1.dbf as a new table document . Make sure, that column ID has type
LONG (if not, convert it – make a new LONG Field using Field Calculator).
• Right click on table sample1, Display XY Data …, accept X_(m) and Y_(m) as
coordinate fields, a map will be displayed as an event theme sample1 Events:
Not all functions in ArcGIS work directly with event themes. It is a good advice to convert
the theme into a shapefile and proceed with that: Right click on sample1 Events Data
Export Data … Save as sample2.shp Shapefile.
5.2 Visualise areal characteristics by a THIESSEN map
A very simple regionalisation method is to assign each point in the domain the value of
the closest observation point:
• In the attribute table of Shapefile sample2.shp, make field ID the Primary Display
Field (right click on sample2 Properties).
3. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-3
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
• Spatial Analyst Distance Allocation… produces the desired map. Field Value
in the attribute table holds the ID of the nearest monitoring point.
• Open the attribute table of the THIESSEN-Grid thiessen1 and JOIN with table
sample2.dbf based on fields VALUE (in grid) and ID (in sample1.dbf) respectively.
Now each pixel carries all attributes of the monitoring points in sample2.dbf and we
can use each of them for display. As an example, we display the electric
conductivities Lf.
4. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-4
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
5.3 Deterministic Interpolation: IDW and Splines
• Default is IDW interpolation using power 2 bull’s eye effect (points appear as
being local extremes contours are like rings = „bulls’s eyes“ around them)
• IDW with power 12: approximates THIESSEN (a site dominates ist neighbourhood).
5. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-5
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
• IDW with power 0.1: approximates global mean; only at the sites the measured value
is reproduced exactly (because a site is only by pure chance exactly at the center
point of a pixel, the result very often looks like an extremely smoothed surface!)
• Interpolation with regularised splines Over- and undershoots!
5.4 Geostatistical methods: Ordinary and Universal Kriging
• ArcGIS Spatial Analyst has only limited support for Ordinary and Universal Kriging
interpolation. In particular, there is no support at all for semivariogram analysis.
• Therefore we use Geostatistical Analyst. It offers a comprehensive list of options in
Explore Data, which help to analyse the data visually and statistically to check for
the assumptions necessary for the application of various interpolation methods.
6. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-6
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
• „Geostatistical Wizard“ is a powerful tool to guide the user step-by-step through
geostatistical interpolation, from analysis via cross vaildation to interpolation.
• Results: estimated vaules and variance
7. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-7
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
• Exercise: Improve the interpolation by adjusting the search neighbourhood
• Further exercise: Use Universal Kriging
5.5 Comparison of methods
• There are many options to compare interpolation results, with cross validation (in
Geostatistical Wizard) being the preferred choice.
8. TUTORIAL: INTERPOLATION OF HYDROLOGICAL VARIABLES T1-8
GIS IN HYDROLOGY AND WATER MANAGEMENT 2010
• Statistics of field Lf: Open attribute table of sample2. Select field Lf. Right click
Statistics
• Statistics of a grid: GA produces only a temporary preview of a grid. To make the
grid permanent (and accessible for other tools) right click Data Export to
Raster. A click on generates a histogram. Further, GA has a direct comparison
option for 2 grids: