Parametric and non-parametric correlation, a time-series cointegration analysis, and possibly a Fourier or wavelet decomposition, if you wish.

Hi Tawhid Hossain,

in terms of the tool box of statistical methods, a simple correlation analysis should do the trick in your case. By performing it, you'll see whether your N and t series are statistically correlated in any way. However, if you wish to go further and probe into potential causal connections between those two parameters, context-specific meteorological/climatological methods would have to be applied. Not sure how long your time series is and to which spatial extent/scale it pertains.

Hope this helps.

Best,

Julius

Thanks Rolando Gonzales Martinez and Julius Riese for the answers. I will look at the methods now before going for a specific method. would you suggest any specific tool or software or tutorial link, please?

Md Tawhid Hossain , I would suggest to use R, Python and -particularly for your needs- MatLab. But a simple tool as Excel can do the trick if your statistical analysis is not too complicated.

https://www.mathworks.com/videos/downloading-and-analyzing-weather-data-68782.html

I believe that a cross-correlation function is the most appropriate tool for looking into interrelationship of the two time serieses. A cross-correlation function shows the correlation coefficient of the corresponding time series at consecutive time values, i.e a function at discrete points of time (hour, day, week, month, etc).

I attach, as an example, a simple Excel spreadsheet. In it, column A contains time lag units, that could be an hour or a day, or week, etc. Columns B is data1 (say, temperature) and column C is data2 time series (say, precipitation). Column E is the cross-correlation function of data1 and data2.

As you can see, data1 and data2 look uncorrelated. However the cross-correlation shows a strong correlation shifted every 3 time units lags at 2, 5, 8, 11, etc

@Tawhid; Correlation coefficient (R) will be helpful when considering the degree of relationship between the two variables however coefficient of determination R^2 could be helpful when considering how the variation in one is explained by the other.