decompose timeseries including NAs

Question

I have a data set including timestamps and (air temperature) numeric values (this including NAs). The time series is in resolution of 10 minutes and covering a period of several years. I would like to do a decomposition of the timeseries. I assume that my data have several cycles:

1. trend
2. seasonal component (e.g. cycle within a year)
3. shorter cycle (e.g. multiweekly to multiday cycle)
4. periodicity (e.g. daily cyle)
5. random noise

I would like to extract at least 1 and 2 and best also 3 and 4 from this data. I already tried with decompose () and stl (), both did not work on my data, due to the NAs.

> dput(head(temp))
structure(list(TS = structure(c(874659600, 874660200, 874660800, 
874661400, 874662000, 874662600), class = c("POSIXct", "POSIXt"
), tzone = "UTC"), value = c(2.77268641140869, 2.77268641140869, 
2.77268641140869, 2.77268641140869, 2.71474135464008, 2.71474135464008
)), row.names = c(NA, 6L), class = "data.frame")