What CAEBM can teach us by analyzing 10.000 years of Holocene:

Some info about #Earth_Warming, gained from #Holocene data by #"Experience_Based_Modeling"

Rudel Stricker

rst.tbus@gmail.com 

In a Nutshell:

Deploying #Experience_Based_Modeling (EBM, c.f. Appendix 2, [01]) , the Earth Warming (EW) "Experience" from 10.000 years of Holocene is used, to identify some details about: "What's going on in terms of EW now, and what can be expected in the near future?" The results indicate, that todays #climate_measures are overdone and mostly worthless. A path to efficient “climate policy” comes into sight.

What #Holocene data told us:

1. We found it possible to construct (c.f. 11,12) a dynamic(!) EBM  "dt.p100.EBM.2020",  capable of predicting over 100 years the #temperature_anomalies (dt)  for all the 10.000 years of Holocene with an ==> RMSE less than  0.05 degC.

Ergo: 100 years predictability means, that all our expectations from short term  #"climate_measures" are unrealistic, because there is a time lag of at least 100 years between #earth_warming and its reasons.

2. Based on Holocene data, extended by measurements up to 2020, the EBM predicts for 2120  ==> dt.2120 = 1.1 degC, while dt seemingly approaches a local maximum then. (Detailed evaluation can be done.) 

Ergo: This level of EW for 2120 is much more lower than the predictions of todays #climate_models, based on (incomplete!) physical laws!

3. TSI, CO2, and dt itself were implemented in the EBM as explicit _dynamic_ parameters: Results show CO2 to deliver the biggest overall contribution to dt, while the influence of TSI is much smaller.

Ergo: The priority of CO2 reduction shows up, in accordance with todays science.

4. As until 1820 there was not so much a big #"men-made_EW" at all, a model "dt.p100.EBM.1920" (based on examples up to 1820) can show something like the #"natural_part_of_EW", e.g. ==> nat.dt.2020 = 0.3 degC, leaving space for the influenceable (= men-made) part of EW as ==> mm.dt.2020 = 0.6 degC.

Ergo: The men-made EW is considerably smaller than the total EW, reducing the potential of any #climate_measures, when based on CO2.

5. As CO2 is implemented as an explicit parameter, the EBM can also support related "What-If" questions, like e.g. What dt reduction could be expected in 2070, if we e.g. had kept CO2 at a constant level, starting e.g. in 1920 or 1970? The EBM's answers are: ==> dt.red.2070.red.1920 = -0.1 deg, ==> dt.red.2070.red.1970 = 0.0 degC.

Ergo: Todays expectations about results of CO2 #"climate measures" are heavily overrated!

6. Disclaimer 

The results shown here are not our "personal opinion", but rather some "infos”, that EBMs can gain from 10.000 years of "EW- Experience" during Holocene. Of course, the correctness of these results is influenced by the quality and correctness of the Holocene data, used to construct the EBMs.

Any questions? 

Comments welcome!

Appendix 1: What we did:

11. EBM setup for dt prediction:

We used EBM (c.f. Appendix 2, [01]) to construct a 100 years prediction model prototype for the temperature anomality dt during the Holocene. As we concentrated on the _method_ at the first run, we used some "preliminary" (unreliable?) data digitized from printed diagrams.

12. Explicit parameters and data update:

In addition to the (EBM-inherent [01]) implicit representation of all influences on dt, we included some influence parameters (TSI, CO2, and CH4) explicitly, to be ready for e.g. some "What_If" studies on them.

We also refined our ("Experience"-) data base for the Holocene by digging out some digital data from public sources, c.f. [02] to [04].

Appendix 2: About #(CA)EBM:

#"Experience_Based_Modeling" (#EBM) [01] is as old as mankind. But due to the tremendous development of science over the last centuries, and additionally because of the increase of computer power lately, #"Science_ Based_Modeling" (#SBM) is used today nearly exclusively in any application area.

But there are still some e.g. highly interdisciplinary, only partially science-covered, and\or (too) complex problems, where #EBM can show it's advantages, e.g. if it comes to identification of even unknown “patterns”, especially if EBM is "computerized" appropriately. And of course in harmony with #SBM, e.g. used as a "Source_of_Experience"...

Therefore: Based on our work with #CAE and #"Computer Aided EBM" (#CAEBM) , we used a bit of the "EW- Experience" from 10.000 years of Holocene, to dig out some info on: "What's going on with EW now, and what can be expected in the near future? "... Of course, there is potential for more…

References:

[01] About (CA)EBM: https://www. tbus-world.com/caebm-work/

[02] Kaufman, D.S., Broadman, E. (2023): Revisiting the Holocene global temperature conundrum. Nature 614, 425–435

[03] Steinhilber, F., J. Beer, and C. Frohlich (2009): Total solar irradiance during the Holocene, Geophys. Res. Lett., 36, L19704

[04] Koehler, P.; Nehrbass-Ahles, C. et al (2017): Compilations and splined-smoothed calculations of continuous records of the atmospheric greenhouse gases ... PANGAEA

Tools and techniques used:

[05] A computer program was setup in Python: https://www.python.org/about/

[06] For data handling etc TensorFlow was used: https://www.tensorflow.org

[07] For setup and handling of Neural Nets Keras was used: https://keras.io

[08] For optimisation of models by Genetic Algorithm PyGad was used: 

https://pygad.readthedocs.io

[09] All calculations were done on a Pixel 7 Mobile Phone

Please, drop a note, if you are engaged in "Earthwarming' or if you are a"decision maker" in this area, and if you want more info or our suggestions, including a free introductional consulting on "How to get a CABM project down to practical work".

continue