GRID tools

This is a presentation of those grid - methods and tools which are used in producing different competive indicators of regional development. There are also URL linkages to thematic map collection in the site of Regional Council of Päijät-Häme.

1. Grid method in spatial planning
2. Data handling
3. Tools

Grid method in spatial planning

Spatial planning and development is concerning ever largening areas. Accordingly the influencing phenomena become more complex and data less accurate and less comparable. Still there is a need of good indicators that have discriminatory power to reveal differencies and explanatory power to predict area development and policy relevance.
In order to handle different types of datasets in analyses an uniform geographic presentation is useful. Sometimes data is too accurate and tedious to compute that it should be aggregated. Sometimes the data is aggregated to units which are far and few or based on administrative boundaries that are not relevant for the topic under examination. Grid method is a practical way to combine data of different type and accuracy.

• Data handling
  • Grid factors can be combined
    To monitor spatial development integrated indicators are needed. With grid method it is possible to combine several factors to look at their added influence.
  • Grid is scalable
    Grid method is practical because the aggregating and disaggregating of data is possible. Vectorbased grid data can be converted to rasterformat and rasterbased data back to vector format (point) data. Same methods are applicable in different scales and geographical scopes.
    • Accessibility potential of BSR centres
      A synthetic database of grid cells sized 10 km were made first. In the BSR area map the accessibility calculations were made by set of grids where each grid cell had an value of accumulated gravity potential of all affecting centres.
      The BSR Map data only consisted the drainage area of Baltic sea, so the accessibility potential preseentations lacks the influent of centres outside the area. So the values are not correct at the edges of the area. For instance Belorussian, Norwegian and German centres are not affecting as they should because of lacking data.
• Data presentation
  Grid method is excellent device for monitoring changes which can be valuable for teritorial differencies.
  • Grid presentation can reveal the similarities and differencies
    Grid presentation is appropriate when studying dimensions of phenomena in which the spatial relations to neighbours or other phenomena are essential.
    • Vague signs of rising factors may be found
  • Catching the phenomena
    The idea of using grid methods in spatial planning is to find out what is happening and why. By detailing different factors spatially it is possible to make presumptions of influent forces and predict the coming.
    • Some indicators describe well the phase of development
      • Average age of population in 1995
      • Taxable income in 1995
      • Change in population in year 1980-1995
    • It is essential to define the influential factors
      • Accessibility of centres
    • Visual comparisons
      • Similar - different areas
      • cause relations with different phenomena
  • Good presentations increase knowledge
    • Interpretation
      Illustrating of different statistics may bring up new ideas depending on earlier knowledge of issue and area. It is also obvious that the interpretation of grid maps varies a lot.
    • Grid maps are impressive in promoting ideas
      Looking at grid maps is like learning by seeing. A subconscious adoption happens when one is looking at a new image concerning familiar area and prosessing the message together with present ideas of it.
    • Lies as any statistics
      In spite of the accurate data and relevant methods used there is always a subjective influence of the author in every grid map. The classification of values can be adjusted and presented so that some values are emphasized. Usually the classifications and color ramps are calibrated so that the results seem reasonable in a well known area. Working with grid maps is often tedious teasing of data until it gives the truth.

Data handling

• Appearence of grid data
  The origin of grid data is usually point data, which is aggragated into grid. When the grid data is stored in database format it is possible to store several attribute fields for one point record and calculations can be made direct with the database.
  • point
    A pair of coordinates is enough for representing the location of an grid cell.
    • x,y,z
      Digital elevation model is usually a set of known x,y,z coordinates strored in a database.
    • Attributes
      Besides location there might be numerous attributes captured from different datasets or calculated by them. Attributes can be everything measurable or logical values. When working with database it is possible to store new attributes for each point for instance traveling time to nearest town over 10 000 habitants or radius in which nearest 500 inhabitants are.
  • Shape
    The shape of an grid is usually square, but any symmetric shape as hexagon or triangle will do.
  • Rasterformat
    Rasterformat comes from image prosessing where the format determines how the location and pixel value is stored. In rasterformat the grid cell has only one value but different attributes can be combined as layers with raster algebra.
• Aggregate
  Decision making is often bound with administrative boundaries and summaries are needed by these areas. Privacy reasons may sometimes force to present gridmaps less accurate. The systematic expression of grid data allows several aggregating methods.
  • Grid size
    Grid size can be enlargen and the attributes summarised to new grid size so that the presentation of phenomen is still valid but the amount of cells becomes reasonable.

    Example map of Päijät-Häme has 1 km grid size

    

    Example map of Finland has 5 km grid size.

  • point in polygon
    Grid data can be aggregated into polygons like admistrative boundaries in the limits of the gridsize.
  • point to grid
    Any point data can be aggregated into grid format with simple spreadsheet or database operations.
  • Area to grid
    Area to grid operation is determing in which area each grid cell belongs and the area code can be then be put as an attribute in grid database.
  • Attributes
    Aggregations can be made by any measurable attribute value and classification that exists.
  • Zones
    Different classifications can be used to produce zones like distant buffers from highways.
• Disaggregating methods
  Spatial analysis packages usually have surface analysis properties. They can interpolate new values for grid points between the original points.
  Hotspot analysis can create raster surfaces using pointdata as origin.
  Accessibility potential analysis is a prosedure where for each gridcell is calculated totaled gravity of all attraction points using their attributes as weighting component and distance function as diminishing the influence of distant attractions.
  Attractions seem to influence differently to their surrounding. Long commuting distances show that workplaces are the main attraction that predicts the urban growth.
  It is obvious that gravity models are explaining lot of actual concentration development.

Tools

• Visual analysis
  The visual analysis is the power of GIS methods. Sometimes visual analysis is also enough. The user can use his or her eyes and brains as analysing tools.
  Basic method to examine spatial data is illustrating it as it is but also after complex analyses the results often are illustrated by classifying or other means.
  • Symbols
    Symbols are efficient way to show where particularly phenomen exists.
  • Classes
    Usual way to illustrate grid data is color ramps. Finding natural colors to represent the phenomen is laborious but essential.
  • Circle
    When the data is gathered from different sources and the quality and accuracy information is missing the visual examination may reveal some quolity matters. Map BSR population is differently represented in different countries. In Poland the population data seem to be classified in fixed classes as in Sweden it seems more accurate.
• Network analyses
  Network analyses can be done by vector format and raster format data.
  • Time distance using traffic net attributes
    The traveling time to Helsinki from every 1 km grid cell has been calculated by the road speed attribute in road register. As result the traveling time is stored as attribute data for each cell.
  • Roadnet with speed attribute
    In calculating traveling time between points it is possible to use measured driving speeds which are attributes in Finnish road register. Since there are not all minor roads in the register the average speed to get to the nearest registered road may be given from experience.
  • Traveling time between pair of points
    With AS2 software a traveling time from one point to all target points can be calculated and stored as attribute for each target cell.
    • Inhabitants within one hour
      Knowing the traveling time to a certain point from a grid data including population data it is possible to calculate how many inhabitants can reach it in a certain time.
    • How far is to the 1000th closest pensioner
      Accordingly it is possible to index the grid with data of nearest pensioner and see how far is the 1000th closest one.
  • Traveling time to the nearest of set of attractions
    Usually there are available several attractions of different kind and in different distance.
    By AS2 software it is possible to calculate distance or traveling time to the nearest of a set of attractions and store the result of traveling time and attraction as attribute for each cell.
• Raster analysis
  Raster analysis algorithms are fast method to manipulate raster data, but they cant handle multiple point datasets.
  • Map Algebra
    Dana Tomlin , GIS and Cartographic modeling, 1990
• Accessibility potential
  Accessibility potential is the simultaneously affect of a certain attraction points measured in every grid cell of the dataset. Attraction points can be cities or services of which volume is known and describes the attractivity. Affect is depending on the closeness of the affecting object. Usually the affect weakens to 1/4 as the distance becomes double.
  Accessibility of two set of grids can be calculated using one set as origin and the other set as target. For instance if the two set are services and customers. The accessibility potential of services presents the situation of customers side showing where is the best service level. Using customers as target the result would present customer potential in every service point.
  

  Accessibility potential is the totalled influence of services divided by the distance in square.
  If we store the accessibility potential value of different targets to a base set of grid it is easy to analyse the influence of several simultaneously affecting components.

  • Traveling time between points in two datasets
    When calculating the traveling time from all points in one set of points to a set of other points the pairs of points are so numerous that the traveling times are not stored but just the accessibility potential.
  • Formula
    • Ai=sum(Pj*f(dij))
      • Ai= Accessibility potential in point i
        The totalled acessibility of each attraction j affecting in point i
      • Pj=attraction in point j
        Attraction is for instance amount of population or services.
      • dij=distance between points i and j
      • f(dij)=dij(exp-2)
  • Usability of accesibility values
    By accessibility calculations we can describe the influence of variables that are continuous rather than discrete.
    • Generalization
      Calculating accesibility potential using the same dataset as target and attraction produces a generalization of the distribution.
    • Accessibility potential
      • Accumulated potential for business, innovations
      • Pressure for land use changes
      • Potential of customers to use the area for recreation
    • Calculations on propability of situtation decisions
      • Grid Arendal population density maps
  • Reference
    A wide review of accessability in theoretical framework and presenting existing indicators is published in
    Study Progamme on European Spatial Planning,
    Working Group 1.1,
    GEOGRAPHICAL POSITION, Final Report, PART I,
    Prepared by the National Focal Points of Finland, France and Germany, Draft, 12 January 2000
    

    SPESP site in Nordregio

Top of page

Send comments to : erkki.rope@paijat-hame.fi
Last update: 29.12.2000