using pre/post datasets for interpolation

[gianguyen]

Hi all, 

I downloaded the daily soil moisture product, AMSR2/GCOM-W1, from

NASA. I attached here some of them, which I already converted and clipped them. 

Some of them contain too many holes or missing data. The missing data is so too big that FocalStatistics doesn't work well. Or this method is not an appropriate one.

I want to use previous/post daily data to fill the missing data for the day in between. For example, if data on day 16th is missing, I want to use data on day 15th and day 17th to interpolate/calculate data day 16th. Could you please guide me to solve the issue?

Thanks

[gianguyen]

I found a keyword for this issue 

Moving Average. 

There is a post "Tool for a moving average over a time series" in 2015, mentioned this method. However, it was used to smooth the series by average them out. I don't want that. I just want to fill the missing data. 

Hopefully, someone can help me!