Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have XY coordinate data of water depth (n = 30) where each coordinate has water depth attribute data for multiple dates from 2000 to 2007. I have imported the data in r for statistical analysis. I want to analyze water depth across different time periods of the year. My dates are as column headings, for example: coordinates, date1, date2, date3,date4 etc and the depth is under date column headings. What statistical methods should I use to analyze and plot the information?

share|improve this question
Have you looked at the spacetime package? – Spacedman Sep 8 '12 at 15:53
What do you mean by "analyze"? Some possible interpretations of this would include interpolating the data in space, interpolating in time, making predictions, summarizing values, running models, fitting trends, etc. Could you also explain what your question has to do with ArcGIS or even GIS at all? – whuber Sep 8 '12 at 18:44
Hi @Shiuli - welcome to the site! Mapping the points is fairly straightforward, but what kinds of analyses do you want to do on the depth data? – Simbamangu Sep 10 '12 at 5:29
I have to echo @whuber. Temporal analysis is an entire field of statistics and some idea as to your specific question would allow us to provide more relevant recommendations. Addressing spatial/temporal dynamics is a real can of worms. From a regression standpoint you could use a STARMA model and, as pointed out previously, there are some options for interpolation. If you have fixed points in time I would recommend looking at the serial autocorrelation and fitting a trend (e.g., LOWESS) to the data before additional analysis. In this way you can address independence and extreme events. – Jeffrey Evans Dec 4 '12 at 20:51

A better answer is certainly possible if you could provide an example data set or more clarity on the term 'analysis'.

That said, if you already have your data in R, an excellent place to start is with the spatstat package. It has wonderful documentation on how to analyze point data that will give you insight into the full range of options.

If this seems like overkill to you, a relatively simple approach (assuming that the multiple dates are the same for all of your observations) would be to interpolate a surface for each time period being sure to define your surface so that raster cells are consistent (e.g. they line up exactly when overlaid on top of one another) then do an analysis of means, mins, maxes etc for each raster cell.

A quick example of doing interpolation on point data using the meuse data set can be found here:

coordinates(meuse) <- c("x", "y") 
coordinates(meuse.grid) <- c("x", "y") 
gridded(meuse.grid) <- TRUE 
idw.out <- idw(zinc ~ 1, meuse, meuse.grid, idp = 2.5) 
spplot(idw.out, "var1.pred") 

Once you have the rasters for each time period the raster math should be relatively straightforward using the operators from the raster package The rasters you produce from this analysis will give you something to work with at least.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.