I have measured a variable (soil moisture) for multiple samples over time. Soil moisture increases steeply at several points in time (when it rains) and then decreases steadily until the next rain event. I would like to express in one value per sample how similar my samples are in terms of soil moisture over time.
One idea I had was to do PCA and to use the values of PC1 as my similarity measure. However, this does not take into account the time series nature of the data.
Do you have a suggestion on how I could take this into account? I could calculate a similarity matrix from Spearman correlations. However, I do not know how to proceed from there...
I am also open for completely different approaches.
