@Doc Hamdeh Kazemi, thanks for your interesting question. You have referred to long-period climate projection deviation from the actual observed record.

My input is that projections are usually

made with mathematical models developed by human beings. These models are components of several interwoven parameters that eventually determine the climate. These include temperature, humidity, atmospheric pressure and other uncontrollable parameters. As long as human beings cannot fully capture and harness the temporal nature of those parameters, there is likely to be deviations from reality.

Great question dear Hamideh Kazemi ,

I think the deviation in GCMs is important just when we want to use them directly, without any preproccessing. Otherwise, we must try to biascorrect or downscale them using different methods and algorithmes. Therefore no one does not use them with deviation.

Secondy, I think the ability of GCMs in simulating the historical period and their error is not a good criteria for assessing their capability, because they have been utilized to track the impact of global climate change on climate parameters and should not be criticized as same as a satellite (data). In other words, they are bulit for generating the future data based on different scenarios, which are much more complicated than our abilities to implement them without any mathematical model.

to sum up, whether historical periods of GCMs have error with observed data or not, they have a percentage change with their future period data which is worthy, merely. for more clarification, just think about scenario implementaion in LARS_WG. You can assess any scenario impact just by calculating the change factors for precipitation and temperature without any bias correction.

Anyway, in recently introduced datasets, the discussed deviation is highly decreased.

(the basical point of view have been discussed before , I just tried to highlight the importance of this subject in practical insight)

Even "reliable sources" most likely have based their projections out of necessity on a limited number of biased locations of meteorological stations. For an example see:

Article Pinus mugo Krummholz Dynamics During Concomitant Change in P...

Article Where the bears roam in Majella National Park, Italy

Hi Kazemi,

The currently applied observation data, from both land and ocean, still have some systematic bias. For example, land surface air temperature data are affected by urbanization effect, and land precipitation data have a wind under-catch related bias. These biases have not been adjusted in the historical climate datasets currently used for both monitoring/detecting climate change and verifying/validating climate models used for climate projection. The urbanization bias in the historical surface air temperature data series leads to an overestimate of the global and regional warming trends and the sensitivity and future temperature projection of climate models, and the under-catch related bias in precipitation data also causes an overestimate of mid- to high latitude land precipitation increasing trend.

In the last two decades, we made some analyses of the observation (in particular surface air temperature) data systematic biases for China and global land, and we would be happy to discuss this issue if you and other colleagues are interested.

Regards,

Guoyu Ren