This is a repository for the climate networks project started in Jan 2022 under Dr. Amit Apte
-
Climate networks in my opinion don't seem to be well suited to capturing the variations in rainfall pattern over the Indian subcontinent. The project managed to show that the climate networks do not follow a power law distribution which many problems of interest seem to follow, and is quite well studied. Instead an exponentila fit was much stronger over the data
-
There are a large number of nodes that have high degree, more than the power law distribution and exponential distribution. This makes it a fat and heavy tailed distribution. This is quite interesting as it means that "IDK ASK AMIT"
-
The degree distributions don't seem to help in discriminating between different rainfall scenarios. Using the KL divergence as a tool to compare distributions, it is clear that there isn't much difference between the distributions. This is quite obvious visually too.
-
Could there exist other network measures that could capture this kind of information? "DISCUSS WITH AMIT"
-
Is it worth pursuing this further? Because there seems to be some work done on this topic and this is a negative result, so it would be interesting to know where this could be taken from here "ASK AMIT"