how well the modeled data fit to the observed data?

Hi,

i am working on timeseries data set. i have gone through many video lectures for the timeseries analysis. all of them talk about decomposing the Time series data.

then modeling it for future predictions.

i have another scene for my thesis project. i have observed data and the simulated/ modeled data already.

both Time series are hourly stamped to show the water levels of a sea. i need to check how well the modeled data fit to the observed data?

i have been stuck in this issue since many weeks. please can anyone guide me a bit , what should i start with?

i have daily hourly data for 46 years. i have decomposed both timeseries into monthly TS, and yearly TS.

i have converted my data into timetables. the data is attached as a mat file .

Best Regards.

Mrs.Raza

[Akira Agata]

To evaluate how well the modeled data fit to the observed data, RMSE has been widely used.

But looking at your data, there should be an offset between these two data, as shown in the figure.

Is it OK to adjust the offset of these two data before calculating RMSE?

[bushra raza]

Hi,

thanx alot for your reply Sir.

yes, i did not adjust the offsets of the data yet. how can i do this? they are original data sets as i recieved.

i had calculated RMSE, and Nash Sutcliff as well,

i am attaching the main.m file, it is using some of my other functions to find the RMSE and Nashsutcliff.

the results are like :

>> main

RMSE=

497.2241

NSE=

-438.9240

i am not finding a way to interpret it.

Basically, i need to transform the simulated data in a way that it represents the variability and mean of the observation.

like for example,i need to fix the 95% variability and the mean of simulated data to observed data.

and ultimately, i need to represent how well the modeled data fit to the observed data?

but how to do it and how can i interpret the result? please guide me.

Regards,

Mrs.Raza

[bushra raza]

Hi, just use 'detrend' for the observed data set. then by using 'distribution fitter' , the following density fit is plotted.

please guide me, how to interpret it? and like 95% variabilty of simulated data to observed data