Go through the code, see the issue no.8 in the landmodel repo
Time and collar
Rs_annual and collar insertion depth by vegetation types, I used multiple linear regression (with weight by the number of observation), vegetation is significant, but only a significant relationship was found under desert. For the shrubland, the trend is greatly affected by the point of collar depth = 10 (but this point is not an outlier).
Then I separate by RC groups, no trend was found
Rs_annual vs measure interval by biome, seems no difference between different biomes
Estimate global RS estimate bias related to measuring time and collar characteristics
Based on the analysis, only need to analyze the bias related to collar length, collar area?
Surrogates
Jianqiu gave me two really good papers
Read through those papers and try to improve the discussion
Landmodels
Go through the code, see the issue no.8 in the landmodel repo
Time and collar
Rs_annual and collar insertion depth by vegetation types, I used multiple linear regression (with weight by the number of observation), vegetation is significant, but only a significant relationship was found under desert. For the shrubland, the trend is greatly affected by the point of collar depth = 10 (but this point is not an outlier).
Then I separate by RC groups, no trend was found
Rs_annual vs measure interval by biome, seems no difference between different biomes
Estimate global RS estimate bias related to measuring time and collar characteristics
Based on the analysis, only need to analyze the bias related to collar length, collar area?
Surrogates