Question | Answer |
What is spatial statistics? | Spatial statistics is the analysis of characteristics of data across space. Sometimes it is hard to visually spot patterns in the data. Spatial statistics techniques helps identify & answer questions spatial patterns and relationships in the data. |
What factors should you consider when exploring your data? | 1- step: Data Exploration, miss key info if skipped and lead to incorrect conclusions and decisions.
The interaction presents the kinds of questions you should ask about your spatial data before beginning an analysis project. |
What is the spatial/geographic location of your data? | This is answered by exploring the spatial distribution of the data. How large is the study area? Are the locations in the study area close together or far apart?
These are few of the patterns worth exploring to learn more about where the data locate |
What are the most common data values? | This is answered by exploring the distribution of the data values. How often do certain values occur? Are the data values close in value, or is there a large range of values? How do the values compare to each other? |
How is the value of the data related to its location? | This is answered by looking at the relationships distribution and the distribution of data values. When performing spatial analysis, there is an assumption that data values close together are more related to each other. |
How do you use this info to select an analysis tool? | Based on your answers to the previous question, you have gathered enough information about your data to select an appropriate analysis tool. |
Where is your data? | It is important to understand the location of your data relative to how it is distributed across space.
Examining the spatial distribution of your data allows you to quickly see where data is clustered, or closed together relative to the rest of the data |
Data Exploration | Visually inspect the data as a part of the process. This first step can help you quickly assess areas in the data investigation. |
Spatial statistic tools | help you better understand the distribution of your data across space. The mean, or average, center of the data is determined by calculating the average x- and y-coordinates of the dataset. |
An average can be pulled, or affected, by very high and very low values. | It is a good idea to compare the mean center with the median, or middle, centre. This is determine by placing the x- and y-coordinates in the number order, then selecting the value in the middle of the list. If the value is very different than the MC. |
Compare the median center with the directional | This tells the general orientation of your data base on the rotation of the ellipse. The directional distribution also tells you the spread of the data based on the std deviation of the x- and y- |
What are the most common data values | |
Trend analysis | Based on direction and on the order of the line that fits the trend. The trend line is a mathematical function, pr polynomial, that describes the variation in the data. Can be used to compare trend line with patterns in the data. |
The order of a polynomial | is based on the equation used to fit the data. You can determine whether the order of the polynomial fits your data based on the shape created by the line. |
The first-order polynomial | will appear as a straight line |
The second-order polynomial | will appear as an upward or a downward curve (know as a parabola) |
A third-order polymonial | will appear to curve either upward or downward, then curve in the other direction as it progresses |
Interpolation | Creates surfaces based in spatially continuous data. Each surface uses the values and locations of your points to create (or interpolate) the values for the remaining points in the surface. |
Interpolation (cont) | Data is not spatially continuous, but is occurrence (discrete) data instead , should investigate other surface creation techniques, such as density mapping. |
Geostatistical interpolation | Create surfaces using the relationships between the data locations and their values.
Predicts values based on existing data |
Geostatistical interpolation (cont) | Data is not clustered (simple kriging has declustering options)
Data is normally distributed (transformation options are available)
Data is stationary (no local varition)
Data is autocorrelated
Data has no local trends (can remove data/part of prces) |
Global deterministic interpolation | creates surfaces using the existing values at each location. Global deterministic interpolation techniques, use your entire dataset to create your surface. |
Global deterministic interpolation (cont) | Outliers have been removed from the data
Global trends exist in the data |
Local deterministic interpolation (cont) | Use several subsets, or neighborhoods, within your entire dataset to create the different components of your surface.
- Data is normally distributed |
Inverse Distance Weighted interpolation | IDQ is a type of local deterministic interpolation. This technique assumes a different set of characteristics about your data.
- Data is not clustered
- Data is autocorellated |