I want to compare two TEC models. Except the comparison of diurnal vTEC, I want to analyze deviation in mean monthly vTEC by taking single airthematic average for whole month. While doing so, I am averaging quiet day as well as disturbed day TEC. But my study is not adressing the performance of models in disturbed or quiet day. Study is focused on overall performance. So, what I believe is if two models are predicting TEC of same location then both should give the similar average for the month. But I want to confirm, wheather other statistical factor effects the value while mixing quiet day and disturbed day TEC or not?? Is this concept significant in comparing two TEC models?
In my opinion, it's better to distinguish the study into disturbed and quiet period to validate the performance of the model. This is because, the average value of the entire month sometime fails to address the storm effect on the vTEC. By dividing the study into two separate categories, we will have better overview of their overall performance.
You can try the Savitzky–Golay technique to achieve a much better fit (mean/average). For reference, you can check Sasmal et al, 2021.
I do understand that you want to compare the two models. For that, if you only use a monthly average of vTEC you will be missing how each model addresses disturbed periods. The comparisons of the performance when considering quiet periods and/or disturbed periods can give you a better insight on how every model behaves under different ionospheric conditions. Not only that, daily vTEC intercomparisons may result more valuable for the comparison that the monthly average.
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory
- Control Systems
- Control Systems Engineering