Lecture 06 : Raster Data Overview (link to Powerpoint file)

Lab Exercise 06: Raster Data Types

Lab Exercise 07: Rasters and Georeferencing

Homework 04: Vector and Raster Mapping

Homework 05: Georeferencing and Digitizing

GIS in Archaeology: Raster Data Overview

Raster Data Organization

Raster data usually follows a simple array based structure. Data is stored in a series of rows and columns. Each cell is the same size as every other cell. X and Y spacing may differ, but the values will be the same for all cells. Data can be stored as integer or floating data values. Data is normally stored in a projected coordinate system. Geographic positioning is based on the coordinate of the origin cell and cell size.

Cell Size, Accuracy and Efficiency

Critical choice for any raster data source is the choice of an appropriate cell size. Small cells provide greater data resolution but higher data storage and analysis costs. Larger cells create smaller files but there can be a loss of necessary resolution. In general the ideal is for cell dimensions to be 1/2 the size of the resolution required. Raster data sources are not able to map the edges of bounded spatial phenomenon as well as vector data sources, but they are better at mapping data which gradually varies over space.

Major Raster Data Types

Satallite imagery data comes from public and private sources and comes in both Monochromatic and Multispectral formats. Some new providers supply data at less than 5 meter pixel resolutions. Satallite data are an excellent way to obtain quality recent data for a large region. Digital aerial photographs are similar to monochromatic satellite imagery but source data comes from a plane. Imagery is usually at a less then 5 meter pixel resolution (1 and 2 meter resolution is very common). Scanned map imagery. Data elevation models (DEMs) which can be either simple elevation raster coverages interpolated from existing topographic paper maps or created by remote sensing technologies (e.g. LIDAR).

Uses of Raster Data Sets

They are often used as background data for mapping other spatial phenomenon. They can also be used to create semi-realistic visual models. Raster overly procedure can be used to identify suitable locations for other spatial phenomenon. Raster data can be useful in modeling the distribution of interpreted data sets.

DEMs and Contour Mapping

DEMs are an excellent way for storing elevation data but they can be somewhat hard to interpret visually. In some cases it is useful to create a contour map from the raw elevation data.

DEMs and other derived data

DEMs can be used to derive other basic important data as well. The slope of a plot of land is often as important as the absolute elevation. Aspect refers tot eh prevailing compass direction of the slope.

DEMs and Hillshading

A common visual application of DEM data is the creation of psuedo 3-D backgrounds.

Raster Overlay

As with vector data sets, it is possible to combine different raster data sets as well. In raster models overlyaing is known as map algebra. In map algebra, some basic operations are recode, algebraic combinations and boolean combinations. Recoding is a process of reclassifying raster data. Algebraic combinations allow you to add, mulitiply, subtract, etc. a combination of raster layers to produce a new output data set. Boolean operations are used for binary raster data sets (data that is classified as either a 1 or a 0).

Recoding Raster Data

Data simplification through reclassifying can be useful for understanding the underlying data distribution. Data can be recoded into interval ranges. Data can be recoded in to a set of binary responses.

Algebraic Operations

Raster data can be combined by the use of various arithmetic operations.

Map Algebra and No Data

Normal action is that if any of the source data cells has not data then the output of the algebra will also have no data in that cell.

Boolean Operations

Boolean functions are used to combine grids to produce yes/no (or true/false) types of results. There are 4 basic boolean functions: not, and, or, and exclusive or (Xor).