Question about Typhoon Track Simulation Accuracy at Different Resolution in MPAS

[ZIZU296]

Hello everyone,
I am running MPAS simulations to study typhoon tracks using different horizontal resolutions. I noticed something unexpected: When using quasi-uniform meshes with 15km and 24km resolution, the simulated typhoon tracks are much closer to the best track observations. However, when using a 60-3km variable resolution mesh, the simulated typhoon track deviates more from the observations.
For reference, I used the mesoscale reference physics suite and compared the MSLP fields. The result shows that the medium-resolution quasi-uniform meshes (15 km / 24 km) actually outperform the high-resolution variable-resolution mesh (60–3 km) in track prediction. I attached the typhoon track comparison plots here, and the blue line represents the observed values, while the red line represents the simulated values.
So I wonder to ask:
Why would medium-resolution quasi-uniform meshes produce better typhoon tracks than high-resolution variable-resolution meshes? Or could it be that it's due to the lack of coupling with the ocean model? Is MPAS suitable for simulating typhoons?

 

[Ming Chen]

Hi,

Thank you for posting the results.

For typhoon simulations, high-resolution often gives more accurate typhoon intensity simulation, whereas its track simulation is not necessarily better.

Coarse resolution simulation can yield nice track simulation, but the typhoon intensity cannot be well reproduced.

The typhoon track from model results is also affected by various diagnostic methods.

 

[ZIZU296]

Hi,

Thank you for posting the results.

For typhoon simulations, high-resolution often gives more accurate typhoon intensity simulation, whereas its track simulation is not necessarily better.

Coarse resolution simulation can yield nice track simulation, but the typhoon intensity cannot be well reproduced.

The typhoon track from model results is also affected by various diagnostic methods.

Thank you for your response!
In theory, high-resolution meshes should be able to capture the wind speed or the lowest central pressure of a typhoon more accurately, and the simulated path should also be closer to the observed values. So I am quite puzzled by the results I obtained. Are there any methods to improve the simulation of the typhoon's path using high-resolution meshes?

 

[Ming Chen]

Typhoon track is greatly affected by the steering flow, which is more related to the forcing data. A high-quality and fine-resolution input data may be helpful for improving the simulation. This is just my personal opinion. Please refer to the literature for more information.

 

[Hans]

Thank you for your response!
In theory, high-resolution meshes should be able to capture the wind speed or the lowest central pressure of a typhoon more accurately, and the simulated path should also be closer to the observed values. So I am quite puzzled by the results I obtained. Are there any methods to improve the simulation of the typhoon's path using high-resolution meshes?

I think this research is very relevant to your question： Redirecting

 

[ZIZU296]

Typhoon track is greatly affected by the steering flow, which is more related to the forcing data. A high-quality and fine-resolution input data may be helpful for improving the simulation. This is just my personal opinion. Please refer to the literature for more information.

OK, thank you for your suggestion！

 

[ZIZU296]

I think this research is very relevant to your question： Redirecting

Thank you very much for sharing. I will take a good look at this article.