the electrical power power provided at SC and OP circuit is almos zero it means that all power coming from solar radiation that enters the PV module is converted into heat. I have done some numerical analysis and experiments that evaluate these differences see the following references

[1] Electrical and thermal model for PV module temperature evaluation

GM Tina, S Scrofani Electrotechnical Conference, 2008. MELECON 2008

[2] Experimental verification of thermal behaviour of photovoltaic modules

GM Tina, R Abate Electrotechnical Conference, 2008. MELECON 2008. .

A cell at open circuit will emit more energy as luminescence than a short circuited cell. In cells of high quality more energy will therefore be dissipated as heat in short circuit.

SC is therefore my answer to your question.

For more productive discussion there are some photos of experimental stend on the roof of our university. The digital tempereture sensor DS1820 are located on the back- side, in the middle of panel. The sensor are in thermal contact with panel and thermal isolation for environment. Another sensors (at shadow place) are measure environment (air) temperature.

I cannot get this publications. I am not IEE member :-(

[1] Electrical and thermal model for PV module temperature evaluation

GM Tina, S Scrofani Electrotechnical Conference, 2008. MELECON 2008

[2] Experimental verification of thermal behaviour of photovoltaic modules

GM Tina, R Abate Electrotechnical Conference, 2008. MELECON 2008. .

Here is some notes from Conference Papers

Maximum heat dissipation in the module will occur when all the incident solar power stays in the module. In MPP that is not the case, becasue the MPP tracker sinks the power out of the module. SC and OC are more difficult. In OC some power is radiated out of a module as a result of luminescence (radiative recombination), in SC some power is wasted as heat on electrical wires. Both power losses are very low  and it is difficult to estimate which one is higher. There is something else that you have to be careful about, in SC there will be some significant differences in temperature between different cells in the module.

While I agree with all answers, it is important to remember that the solar radiation falling on the solar module does not change in value, when you are measuring the temperature of the cell under different operational conditions. Moreover, it is important to appreciate that the temperature read is actually lagging behind the actual change in operational conditions, due to the thermal mass of the module itself. It is therefore important to carry out such tests under steady state conditions in the labortaory. However, theoretically speaking the question has already been answered.

the positioning of the sensor is also important there is a gradient from the center to the border of the PV module and of course near the junction box.

Also the presence of frame influences the gradient.

Thank's to All!

And what do You think about heat transport from front-side to back-side of PV-panel? Something like Thermoelectric (Peltier) effect?

Let us to continue discussion with may question "What are the factors which limite overheating of single cell in PV-battery?"

https://www.researchgate.net/post/What_are_the_factors_which_limite_overheating_of_single_cell_in_PV-battery

Article PV-battery Power and Heat Balance

Dear Vladimir,

I gave an answer on a similar question by you previously on the research gate. If i understood you, you do not find a strong correlation between the solar cell loading and the temperature at the back of your PV modules. The above answers agree with my comments. If there is any thermal effects for loading , and there is sure effects, one has to measure the cell temperatures directly not the back temperatures. Therefore,i would propose that you may better repeat the experiments with direct measurement of the cell temperatures.

Best wishes

By  Abdelhalim Zekry "may better repeat the experiments with direct measurement of the cell temperatures".

It is impossible due to cell shading by temperature sensor. Then shading leads to cell's temperature incriasing

Dear Vladimir,

You may use contactless temperature sensors like infrared detectors. They may be of help to you. You may include also miniature thermistors during the encapsulation process. I think new experiments are needed to decide about your phenomenon.

Best wishes

By Abdelhalim Zekry "You may use contactless temperature sensors like infrared detectors". This proposition is true. One year ago I had go by this way and I had detected essential temperature anomalies on the PV-battery surface. By accident, one of  the temperature sensors has located in the middle of Hot Spot. This accident confused my previous experemental results on comparative study of two PV-panels in Short Circuit and Open Sircuit modes. So, by last step I had use averaging on the whole surface of PV-panels.

Today I am sure that averege PV-battery surface temperature in Short Circuit mode is equal to surface temperature in Open Sircuit mode.

Thanks to all who help me to find correct solution of my problem.