I have two broad questions.
1) If you wanted to improve daily weather prediction anywhere in the world, what is the low hanging fruit in terms of additional observations you would look to pull in (considering impact on forecast skill and likelihood of implementation) ? Is it higher spatial density and more frequent raobs? or dual-polarized radar ? or ground 2-10m meteorological networks? anything from geo-stationary satellites or polar-oribiters? or simply a higher resolution AOGCM? or dedicated airborne missions (if so, what variables would you measure)
2) Now, the same question for seasonal time scales. What is the area that meteorologists generally point to as a weakness? Anything that can be added to better monitor/predict atmospheric rivers? is the in-situ/satellite SST measurement sufficient?
You can be as brief or in-depth in response as you wish. I focus on remote sensing of the land surface for hydrology and I'm trying to get a sense of the value of satellite observations I use to weather forecasting.
1. Probably more raobs. I've noticed in practice, it's somewhat difficult to infer upper air data in some places. More raob stations would definitely help. Dual pol radar and ground based stations are usually sufficient. Satellite wise, I use visible, infrared, and water vapor in my practice.
2. I'd think water vapor imagery would be best for this. In fact, I have a book that's all about water vapor imagery interpretation. I can give you the title and author if interested.
Hope this helps.