I have meteorological daily sunshine duration, humidity, rainfall, and max-min temperature; 3 hourly cloud cover, visibility, wind speed with direction; global radiation data at 1 minute interval for a few years. I worked on predication of sunshine duration, solar radiation, etc. from other meteorological parameters. Now I have been studying the solar radiation forecasting methods for the last couple of weeks. There is a good number of options, like Clear sky models, regression methods, artificial intelligence, remote sensing, numerical weather prediction, local sensing, etc. in different journal papers. One of fellows are working on DNI measurement, finding the Linke Turbidity Coefficient and using clear sky models for solar radiation prediction for clear days. But I am not sure yet whether I can make forecast of solar radiation for any forecast horizon for Bangladesh. Would you mind to cooperate me in this regard?
Do you only want to forecast the solar radiation or do you want to forecast the whole meteorological situation?
Attention: Mr. Bauer,
Thanks. I am only interested in solar radiation forecasting.
Okay. I suppose that the solar radiation has a seasonal pattern? If it holds, you could use time series decomposition (for instance STL) for getting the trend, season and noise part. Based on this information, you could use to train a machine learning algorithm.
Dear Bauer, Let me check first. If needed I will bother you again. By this time would you mind to add me in your Linked in site?
I'm sorry, I don't have LinkedIn. Nevertheless, feel free if you have any questions :)
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing