|
Dear All,
I have an area (5 ha) split up in sub areas (for example 45, ~ equally spaced). 1/3 of the sub areas (for example 15) have data points and I am interested to predict values for the remaining 2/3 of the area. Each 3rd of the sub areas got a different treatment.
2 Scenarios:
1) A lot of data points (like from on-line sensors, 2-4 m spaced, lets say 2000 points) exist in 1/3 and non in the other 2/3. My first question would be, if that sort of scenario is what’s called a nested data structure? And in case I am using a kriging procedure to predict the rest 2/3 of the area, what would be an appropriate way of doing so? How do I handle the unevenly spaced data? Risks? Concerns? 2) A limited number (let’s say 100 points) exist in 1/3 of the area, spread out as equal as possible (beside of some close data points for a better nugget prediction). So overall, the data points are more evenly spaced, but I have much less data.
I am especially suspicious about the quality of prediction for the remaining 2/3 of the area, because I will compare the 3 different treatment maps spatially and make estimations about the results a certain treatment would have had in the other 2/3 of the area.
I hope I could outline the problem detailed enough, otherwise I can deliver more information. Any suggestions, ideas and thoughts concerning this type of data and data handling are very much appreciated!
Thanks for your help and effort, Cheers, Kerstin
Kerstin Panten CSIRO Land and Water Adelaide Laboratories Waite Road Postal: PMB No. 2 Urrbrae Glen Osmond South Australia SA 5064
Tel. 08 8303 8595 Fax. 08 8303 8550
|
* By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm )
* To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
