Veda2.0 Released!


TIAM - Climate module
#1
Hello everyone, 

We are running a baseline scenario of the TIAM model with climate module so that after we can run and compare our results with a 1.5 degree scenario. 

However, we noticed that the temperature increase for a BAU is quite low (Delta Atm = 1.13 max) whereas GHG emissions are drastically rising. Please see the attached images. 

In your esteemed opinion, what could be the reason? 

Many thanks for your help. 

Best regards,

Sophie


Attached Files Thumbnail(s)
       
Reply
#2
Your global GHG emissions seem to have strange values: For 2018 your screenshot shows the value 5008992 for GHG emissions, but in what units is that?  If it is in kt(CO2 eq.), it is way too small, and if it were in Mt, it would be far too high.  The 2019 total GHG emissions were about 52 Gt(CO2 eq.), from the "Kyoto gases" CO2, CH4, N2O and F-gases, and so your numbers look peculiar.

Another reason could be missing input data for the climate module.  See the documentation, Part II, Appendix A.

Perhaps most importantly, check these parameters:
CM_GHGMAP(r,c,cg)  –  Conversion factors from regional GHG commodities ( c) to global emissions (cg) in the Climate Module
CM_HISTORY(y,cm_var) – Historical calibration values at years y, for cm_var
CM_LINFOR(y,cm_var,lim) – Linearized forcing functions (if you define bounds on forcing or temperature)

If your global GHG emissions are indeed on a correct level (>50 Gt/a), it seems your model may be missing the CM_GHGMAP(r,c,cg) parameters, which are mandatory. For CM_HISTORY, TIMES has default values up to 2010 only, and so one should better add some values for more recent year(s), e.g. the 2020 concentrations, and DELTA-ATM(2020)=0.98 (according to NOAA).

[Edit:] Actually the DELTA-ATM should refer to ΔT above the pre-industrial (1850–1900) average. The provisional 20-year average for the period 2002–2021, has been reported as 1.01 °C above the 1850–1900 average, and so that estimate may be considered suitable as the value for CM_HISTORY(2020,DELTA-ATM).  For the concerntrations, NOAA reports the annual average values in ppm/ppb, which can be used (converting to Gt/Mt and subtracting natural concentrations).


                 CO2(GtC)   CH4(Mt)    N2O(Mt)      ΔT
2010             826           3121         415          0.8
2015             850           3221         454          0.9
2020             877           3349         492          1.01
Reply
#3
(26-01-2022, 01:12 AM)Antti-L Wrote: Your global GHG emissions seem to have strange values: For 2018 your screenshot shows the value 5008992 for GHG emissions, but in what units is that?  If it is in kt(CO2 eq.), it is way too small, and if it were in Mt, it would be far too high.  The 2019 total GHG emissions were about 52 Gt(CO2 eq.), from the "Kyoto gases" CO2, CH4, N2O and F-gases, and so your numbers look peculiar.

Another reason could be missing input data for the climate module.  See the documentation, Part II, Appendix A.

Perhaps most importantly, check these parameters:
CM_GHGMAP(r,c,cg)  –  Conversion factors from regional GHG commodities ( c) to global emissions (cg) in the Climate Module
CM_HISTORY(y,cm_var) – Historical calibration values at years y, for cm_var
CM_LINFOR(y,cm_var,lim) – Linearized forcing functions (if you define bounds on forcing or temperature)

If your global GHG emissions are indeed on a correct level (>50 Gt/a), it seems your model may be missing the CM_GHGMAP(r,c,cg) parameters, which are mandatory. For CM_HISTORY, TIMES has default values up to 2010 only, and so one should better add some values for more recent year(s), e.g. the 2020 concentrations, and DELTA-ATM(2020)=0.98 (according to NOAA).

[Edit:] Actually the DELTA-ATM should refer to ΔT above the pre-industrial (1850–1900) average. The provisional 20-year average for the period 2002–2021, has been reported as 1.01 °C above the 1850–1900 average, and so that estimate may be considered suitable as the value for CM_HISTORY(2020,DELTA-ATM).  For the concerntrations, NOAA reports the annual average values in ppm/ppb, which can be used (converting to Gt/Mt and subtracting natural concentrations).


                 CO2(GtC)   CH4(Mt)    N2O(Mt)      ΔT
2010             826           3121         415          0.8
2015             850           3221         454          0.9
2020             877           3349         492          1.01

Dear Antti, 

Thank you for your reply and explanation. 
Is it possible also to update the CO2-LO and UP values and the natural concentrations of CH4 and N2O from 2005 to 2020 ?

Best,
Sophie
Reply
#4
> Is it possible also to update the CO2-LO and UP values and the natural concentrations of CH4 and N2O from 2005 to 2020 ?

As you can see from the documentation, for CH4 and N2O there are only two concentration "boxes", ATM and UP, of which the UP part represents the natural concentration and is assumed to stay constant.  The anthropogenic concentration is represented by the ATM box.  Therefore, also the calibration parameters CM_HISTORY(y,cm_var) for the historical concentration are for these two boxes:

  ● CH4-ATM – anthropogenic CH4 concentration (Mt)
  ● CH4-UP – natural CH4 concentration (Mt)
  ● N2O-ATM – anthropogenic N2O concentration (Mt)
  ● N2O-UP – natural N2O concentration (Mt)

   

As I am sure you know, the lifetimes of methane and N2O depend on the concentration, and therefore, in particular for methane, one should consider using a lifetime longer than the mean atmospheric lifetime. In the above, using the so-called perturbation lifetime is shown, which however may already be somewhat too high for using in this simple model. I have myself earlier used a value of 10.9 years, which seemed to give a good calibration. If you have suggestions about how to choose the appropriate lifetime for methane, please share.

For CO2, sure one can update the CO2 mass (in GtC) in the LO and UP reservoirs (biosphere and deep ocean) if you have data for them for more recent years (2015, 2020).  If you have such data, could you please also share it here?
Reply
#5
(04-02-2022, 06:06 PM)Antti-L Wrote: > Is it possible also to update the CO2-LO and UP values and the natural concentrations of CH4 and N2O from 2005 to 2020 ?

As you can see from the documentation, for CH4 and N2O there are only two concentration "boxes", ATM and UP, of which the UP part represents the natural concentration and is assumed to stay constant.  The anthropogenic concentration is represented by the ATM box.  Therefore, also the calibration parameters CM_HISTORY(y,cm_var) for the historical concentration are for these two boxes:

  ● CH4-ATM – anthropogenic CH4 concentration (Mt)
  ● CH4-UP – natural CH4 concentration (Mt)
  ● N2O-ATM – anthropogenic N2O concentration (Mt)
  ● N2O-UP – natural N2O concentration (Mt)



As I am sure you know, the lifetimes of methane and N2O depend on the concentration, and therefore, in particular for methane, one should consider using a lifetime longer than the mean atmospheric lifetime. In the above, using the so-called perturbation lifetime is shown, which however may already be somewhat too high for using in this simple model. I have myself earlier used a value of 10.9 years, which seemed to give a good calibration. If you have suggestions about how to choose the appropriate lifetime for methane, please share.

For CO2, sure one can update the CO2 mass (in GtC) in the LO and UP reservoirs (biosphere and deep ocean) if you have data for them for more recent years (2015, 2020).  If you have such data, could you please also share it here?

Dear Antti, 

Thank you again. I can share with you what is found in the below link (pages 20, 22 and 26) there are data concerning land and ocean, respectively. 
Do you think they can be useful ? 
https://www.carbone4.com/wp-content/uplo...2020-1.pdf

Best,
Sophie
Reply
#6
Thanks, I had not seen that report.

Well yes, some of the estimates can be useful, but I would also give some "words of warning":

Page 20 is not really that useful, because we have good time-series data on the atmospheric cooncentration elsewhere, e.g. from NOAA. For example, the average concentration was 414.24 ppm in 2020 (NOAA data), which may be used as an appropriate calibration value for TIMES, as the concentration in 2020. From that we would get 882 GtC (in my earlier post I used the 2019 value 877 for being a bit "conservative").

Pages 22 and 26 could potentially be used for calibrating the CO2-UP and CO2-LO reservoirs, but then you may also have to re-calibrate all the transfer parameters between the reservoirs (the PHI* parameters for CO2). In other words, you should be warned not to define the historical carbon stock values of CO2-UP and CO2-LO (by CM_HISTORY) with any data source values that may be inconsistent with the corresponding reservoir transfer parameters (all the current parameters are based on the simple Nordhaus carbon cycle model).

So, I would suggest that the values in this report may be used for TIMES only if you can also make a full re-calibration or at least a validation of the simple three-reservoir carbon cycle model. If you can do such a re-calibration or validation (checking it with a long historical time series data on anthropogenic CO2 emissions), I think it could be a nice contribution for the new version of ETSAP-TIAM.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)