Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 0 views

    Spatial Interpolation 1

    Uploaded by

    Sreepriya
    interpolation

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Spatial Interpolation 1

    Uploaded by

    Sreepriya
    0 views47 pages
    interpolation

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    interpolation

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    0 views47 pages

    Spatial Interpolation 1

    Uploaded by

    Sreepriya
    interpolation

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 47

    Spatial Interpolation

    Prepared by Dr. Sandipan Das


    Assistant Professor

    Symbiosis Institute of Geoinformatics


    Spatial Interpolation
    • Spatial Interpolation is the process of using points
    with known values to estimate values at other
    points.
    Why Use Interpolation?
    • To estimate values where there is no data

    • Visualization of complex point data

    • Analysis of trends over space and time

    • Creation of discrete boundaries between values


    How does it Look
    • Let say we have our ground water pollution samples

    This
    gives
    us
    Requirements of interpolation
    • Interpolation only works where values are
    spatially dependent—that values for nearby
    points tend to be more similar

    • Where values across a landscape are


    geographically independent, interpolation does
    not work
    What is autocorrelation?
    Tobler’s first law of geography:
    • “Everything is related to everything else, but near things are
    more related than distant things.”

    • This is the basic premise behind interpolation, and near points


    generally receive higher weights than far away points.

    • Positive Spatial Autocorrelation (Neighbors are similar)


    Negative Spatial Autocorrelation (Neighbors are dissimilar)
    Interpolation examples
    • Weather
    • Groundwater contamination
    • Pollution
    • Environmental data
    • Public health
    • Crime
    Method of sampling
    Method of sampling is critical for subsequent interpolation
    Interpolation vs. Extrapolation
    • Interpolation is prediction within the range of our data

    • Extrapolation is prediction outside the range of our


    data
    Elements of Spatial Interpolation
    • Spatial Interpolation requires two basic
    inputs:-

    • Control Points
    • Type of Spatial Interpolation
    Control Points
    • Control Points are points with known
    values.

    • Also called known points, control


    points, sampled points, or observations.

    • They provide the data necessary for


    the development of interpolator

    • The number and distribution of control


    points influence the accuracy
    Classification of Spatial Interpolation
    1.Global vs. Local
    2.Exact vs. Inexact
    3.Deterministic vs. Stochastic
    Global vs. Local
    1. Global Interpolation:
    • It uses every known point available to estimate an
    unknown value.

    2. Local Interpolation:
    • It uses a sample of known points to estimate an
    unknown value.

    • It is important to know how many known points to use,


    and how to search for those known points.
    Three search methods for sample points

    (a) (b) (c)

    (a) find the closest points to the point to be estimated,


    (b) find points within a radius, and
    (c) find points within each of the four quadrants.
    Exact vs. Inexact
    1. Exact interpolation
    It predicts a value at the point location that is the same as is
    known value.

    2. Inexact interpolation
    It predicts (or approximate) a value at the point location that
    differ from its known value.
    Deterministic vs. Stochastic
    1. Deterministic interpolation
    • It provides no assessment of errors with
    predicted values.

    2. Stochastic interpolation
    • It offers assessment of prediction errors with
    estimated variance.
    Classification of spatial interpolation methods

    Global Local

    Deterministic Stochastic Deterministic Stochastic

    Trend surface Regression Thiessen (exact) Kriging (exact)


    (inexact) (inexact) Density estimation (inexact)
    Inverse distance weighted (exact)
    Splines (exact)
    Global Methods - Trend Surface Models
    • An inexact method, trend surface analysis approximates
    points with known values with a polynomial equation.

    Example of a global interpolation – linear trend surface for Swiss rainfall data.
    Regression
    • A statistical method used to examine the relationship
    between a variable of interest and one or more
    explanatory variables
    – Strength of the relationship
    – Direction of the relationship

    • Often referred to as Ordinary Least Squares (OLS)


    regression

    • Available in all statistical packages


    Global Methods - Regression Models
    • A regression model relates a dependent variables to a number of
    independent variables in a linear equation (an interpolator), which
    can then be used for predictions or estimation.

    • Local regression analysis uses geographically weighted regression


    analysis.
    Local Methods - Inverse Distance Weighted (IDW)

    • IDW is a local & exact method


    • It assumes that the estimated value of a point is influenced more
    by nearby known points than by those farther away.

    Estimation surface by using inverse distance weighting; Swiss rainfall data


    Local Methods - Spline
    • It uses mathematical function that minimizes overall surface
    curvature, resulting in a smooth surface that passes exactly
    through the input points.

    • Method is suited for gently varying surfaces, such as


    elevation, water table heights, or pollution concentrations.

    Interpolation of a 50-m resolution DEM with complex topography using regularized spline
    Local Methods - Thiessen Polygons
    • Triangulation process is used to generate Thiessen polygons.

    • Triangular Irregular Network (TIN) models for elevation, slope


    and aspect modelling
    Local Methods - Density Estimation
    • It measures cell densities in a raster by using a sample of
    known points.
    • There are simple and kernel density estimation methods.

    Fig: Deer sightings per hectare calculated by the simple density estimation method
    Local Methods - Kriging
    • Kriging considers both the distance and the degree of variation
    between known data points when estimating values in
    unknown areas.

    • Kriging can assess the quality of prediction with estimated


    prediction errors.
    Steps in Kriging
    • Involves several steps
    – Exploratory statistical analysis of data

    – Variogram modeling

    – Creating the surface based on variogram


    Semivariogram
    • Kriging uses the semivariance to measure the spatially
    correlated component

    • Measures the variability of data with respect to spatial


    distribution(quantifies similarity/disimilarity)

    • Analyzes variance between pairs of data points over a range


    of distances
    Model Fitting
    • A semivariogram must be fitted with a mathematical function
    or model so that it can be used for estimating the
    semivariance at any given distance.
    Model Fitting
    • Five mathematical models for fitting semivariograms
    • Two common models for fitting semivariograms: spherical and
    exponential
    Nugget, Range, and Sill
    • The nugget is the semivariance at the distance of 0, representing
    measurement error, or microscale variation, or both.

    • The range is the distance at which the semivariance starts to level


    off.

    • The sill is the semivariance at which the leveling takes place.


    Kriging Results
    • Once the variogram has been developed, it is used to
    estimate distance weights for interpolation

    • Computationally very intensive w/ lots of data points

    • Estimation of the variogram complex


    Kriging output: prediction
    Types of Kriging
    We can classified Kriging in following categories

    • Ordinary Kriging
    • Universal Kriging
    • Co-Kriging
    Ordinary Kriging
    • Assuming the absence of a drift, ordinary
    kriging focuses on the spatially correlated
    component and uses the fitted semivariogram
    directly for interpolation.
    Universal Kriging
    • Universal kriging is used where there is an
    underlying trend beyond the simple spatial
    autocorrelation
    Cokriging
    • Cokriging is an interpolation technique that uses two or
    more (up to four) datasets to refine the predicted values.

    • Co-kriging is used when there is a second variable that is


    strongly correlated with the variable from which we’re
    trying to create a surface, and which is sampled at the same
    set of locations as our variable of interest and at a number
    of additional locations.

    • E.g Better results in precipitation interpolation have been


    reported by including elevation as an additional variable
    Comparison of Local Methods
    • Comparison of local methods is usually based
    on statistical measures

    • Cross validation and validation are two


    common statistical techniques for comparison.
    A Cross-Validation Example
    • Removing one of the n observation points and using the
    remaining n-1 points to predict its value.

    • Error = observed - predicted

    Interpretation

    Smaller RMS – Better


    interpolation
    Validation Example
    • Divides known points into two samples-
    1. One sample for developing the model
    2. Second sample for testing the accuracy of model
    Spatial interpolation using ArcGIS
    Spatial Analyst Tools Geostatistical Analyst Tools

    • Offer same tools for • Offers explore data,


    interpolation geostatistical wizard & create
    subsets
    • Include histogram,
    semivariogram, QQ plot &
    other data analysis
    • Provides cross validation &
    validation
    • Offers large variety of global &
    local interpolations
    Analysis of Variogram
    How Many Neighbors?
    Cross Validation
    IDW vs. Kriging
    • Kriging appears to give a
    more “natural” look to the
    data

    • Kriging avoids the “bulls


    eye” effect

    • Kriging gives us a standard


    error
    Which Interpolation methods to use?
    • It is not always easy to understand how data behaves before
    commencing with the process and therefore it can be difficult
    to know what technique should be used.

    • However, there are some data set that will help determine the
    most appropriate technique.
    Data Type Possible Interpolation
    • Some interpolation techniques can be automatically applied
    to certain data types

    • Elevation – TIN
    • Soil Chemistry - IDW, Kriging
    • Demographic - IDW, Kriging
    Thank You

    You might also like