I am trying to study the correlation between weather and air pollutants data, to understand How weather influence the air pollutants?.
I tried the Spearman correlation for 5 days observations, but for every time interval in the time series I got different correlation !!
Is that possible to have different correlation for different intervals, and using different measures (like hourly concentration and daily average concentration) of the same time series?
If you have different correlations for different intervals, it appears you are missing at least one relevant factor. Do you look at the whole process: emission -> transmission -> immission?
Ingo Riess I am trying to look for the correlation so I can select some features that are important for my model of prediction. Also, to see how weather conditions effect air pollutants in short period of time. for example, if air pollution caused by the O3 in one day, I want to see what caused the high concentration of Ozone in the 5 previous days.
I looked at Temperature , wind speed, rainfall, and all air pollutants o3, pm10, no2, so2, ..
Wedad Alahmadi this paper may help you: Article Relationships between meteorological parameters and criteria...
They divided their time series (12 months) into 4 seasons, and for each one the Spearman-Rank correlation coefficient value was different.
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining