conditionally write tendency terms to output file (only if the index in the big tendency array is >= 1, i.e. used)
-- this replaces the previous function of always initializing the tendency variables in the output file whether they were used or not in the run, resulting in arrays of missing values
the scm_analysis.py script was modified to treat these variables as optional variables in the output; if they're not present for a given output file, they're filled in with zeros (not NaNs or -999) because they are often plotted together with other tendencies, and it is better to see 0 tendencies than -999 or something