• ​While analyzing the graphs, we found that the hail values are not very evenly distributed, and the size prediction of the hailstones was not very accurate. 

• It is because of limitations on data collection.

• There was lack of samples to plot on the graph and derive the correlations with other indices.

• It only takes about 5 to 10 minutes for hailstones to fall on the ground, and it occurs in narrow areas about 1 to 2 kilometers, which means it has very small scale both temporal and spatial. 

• In order to observe this small-scale phenomenon, generally we need to set up a smaller grid of hundreds of meters and conduct a minute-by-minute observation of the time scale to obtain accurate data.

• However, it is difficult to get temporal data in regions where satellite images, weather maps, and instablility indices does not provide the exact time zone when hail occurs.

• Especially for satellite images of large phenomena, the big limitation is that the observed scale is too small to identify the hail phenomenon in the local area.