I used bootstrap compared the mean of measured Rs and modeled results
At the annual time scale, modeled Rs match well with measured Rs
At the monthly timescale, however, modeled Rs is higher than that from the measured Rs
why? overestimate in winter, separate the results by different seasons?
overestimate in the north ~50 degree region?
Time-scale matter? develop a global monthly Rh and Ra data products is important for the model benchmark?
CMIP6 mean?
Site scale comparison
Can we extract any meaningful functional relationships between the relative or absolute magnitude of these controls over Rh fluxes from the observations than can be compared with the models?
Time for warm
There are totally 83 sites longterm (>=5 years) annual Rs measurements;
Each year I simulated a linear regression between annual Rs and years after the experiment was set up;
In the future, only those site with positive air temperature trend (i.e., warm year after year) were selected;
The black solid lines represent significant trend (p<0.1);
The dotted lines mean p>=0.1;
The red line represents the average of all sites;
We can see that annual Rs do not see an increasing trend even though the air temperature increased!
Why? due to acclimation?
But we do not find a different Q10 of 1990s and 2000s.
And Rs vs Tm relationship does not show a difference in 1990s and 2000s.
Time and collar
Revised based on Chris' comments, I believe it is much stronger now
Maybe ESSD is a good target for this manuscript?
Goal for this week
Submit surrogates manuscript
Continue working on the time for warm and landmodel projects
Landmodels
I used bootstrap compared the mean of measured Rs and modeled results
At the annual time scale, modeled Rs match well with measured Rs
At the monthly timescale, however, modeled Rs is higher than that from the measured Rs
why? overestimate in winter, separate the results by different seasons?
overestimate in the north ~50 degree region?
Time-scale matter? develop a global monthly Rh and Ra data products is important for the model benchmark?
CMIP6 mean?
Can we extract any meaningful functional relationships between the relative or absolute magnitude of these controls over Rh fluxes from the observations than can be compared with the models?
Time for warm
Why? due to acclimation?
But we do not find a different Q10 of 1990s and 2000s.
And Rs vs Tm relationship does not show a difference in 1990s and 2000s.
Time and collar
Goal for this week