Jinshi-Ben-Meeting-2019/3/25 #6

[jinshijian]

Updates from last week

Get soil temperature for all 869 records (Done)

• 351 soil respiration are from MGRsD (average of at least 12 months' soil temperature to get an annual mean soil temperature), based on those data, Ts and Tair are linearly related (*Ts = 2.918+0.829Tair**);
• 163 records have less than 12 months' soil temperature data, the missing Ts are estimated based on monthly air temperature and linear relationship between Ts and Tair (above equation and figure);
• There are 67 records I cannot find the soil temperature information, can I based on the Rs_Ts_relationship and Rs_annual get a estimate of the annual mean Ts? Whenever a paper has reported annual mean Ts, I compared the reported annual Ts and estimated annual Ts based on the Rs_Ts_relationship, I found they are well matched (not well match indicate a potential problem, then I go to the manuscript and check). (I only use this approach when a study has multiple records, and compare the estimated annual mean Ts with Tair, if they are well match, the mean Ts of multiple records are used, we can discuss whether we should include those 67 records);
• The rest are directly read from figures, tables, or description in the paper.

Analysis predicted Rs_bahn and reported Rs_annual (Partly Done)

• When more data added, we got a similar result -- Bahn method slightly overestimate Rs_annual when using soil temperature as predicter (but no difference when using air temperature as predictor).

Test whether Bahn method can be used in monthly time scale (Done)

• More error in monthly time scale (especially in May, June, July, Aug, and Sep, but pretty good in cold months. Due to higher temperature in summer and out the valid range of model?).
• Overall, Bahn methold overestimate Rs_monthly even using air temperature as predictor.

Test whether Bahn method can be used in daily time scale (Done)

• Similar conclusion as monthly time scale, Bahn methold overestimate Rs_diurnal.
• But for diurnal timescale, I used a different approach to get measured Rs when soil temperature approch diurnal mean.
• For example, in a site and within a day (follow figure, black line is diurnal Rs, and red line is diurnal Ts), the red dash line is daily mean soil temperature, based on this value I can get when soil temperature reach daily mean, and then Rs in this time period was defined as Rs when soil temperature reach daily mean.

Goal for next week

Make PPT for ESS meeting and write manuscript for Bahn analysis.

[jinshijian]

• Above figure shows the relationship between Rs_annual and Rs_bahn with three different Ts sources.

[bpbond]

Also want to discuss:

• Temporal variability paper