I need explanation on the number of plot to use for kriging for a study area of 6000 ha
hi Mikwa,
if you want to use rigging, it's a bit more complex. You need to also simulate the boundaries = if you do I rigging between a forest point for instance and a crop point, the distance in between is going to be changed in a progressive transition.... You therefore first need to put weights on your location, and then the complexity of your spatial variability will give you indications on the number of points you need.
Hello. The sample size depends on terrain complexity. If your region is flat, you can design a regular grid for sampling. Otherwise, the sampling density should correlate with terrain complexity (eg. higher sample density in more fragmented areas). As for the number of points, there is no general rule, but you could attempt to estimate it based on the confidence interval of the mean, if the statistical distribution of carbon is normal.
Sorry for my first answer. I just realised that the automatic orthographic corrector modified everything I wrote and it doesn't make much sense anymore.... Bloody tablet!
Hi Jean-Fiston,
What is the question you are asking that requires you to estimate the carbon stock in this big of an area? Chances are that you will make a bigger error, by taking shallow samples (because you need many to do a geostatistical analysis) than you would by taking a few deep samples and calculating their average... Take care, Philippe
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India