Output_variables limit of H_CORR and default set of gapfill_met_tier1 = 10

[zejiang-unsw]

Hi Anna,

I have met two issues when using your R package to deal with Fluxnet data. I am totally new to this dataset, and my questions might be very naive. Thanks for your patience and understanding in advance.

1). For H_CORR, What is the Data_min and Data_max (-1000,1000) based on? I have met the case that is beyond this range, e.g., site RU-Fyo data range is [-248.4, 1058.7] and site IT-Col data range is [-462.4, 1391.5]. Can we set the range to e.g. (-1000,1500)?

2). For gapfill_met_tier1, by default it is 10, why it is different from tier2?

Cheers,
Ze Jiang (UNSW, Sydney)

[aukkola]

Hi Ze Jiang,

1. The variable ranges mostly come from the CABLE model. They were designed to catch any outliers, but if you want to change the ranges, you can modify the ranges in data/Output_variables*.csv. Another option is to set the check_range_action to one of the options below to stop the code from aborting:

#' - check_range_action: Action to take when input data falls outside of valid ranges
#'        (as defined in data/Output_variables_*.csv).
#'        One of "stop" (log error and stop processing, default),
#'        "warn" (log error and continue), "ignore" (continue), or
#'        "truncate" (set values outside the valid range to the range bounds).

2. We deemed tier 1 variables to be more important and set stricter limits for those in terms of gapfilling allowed. For example, we were willing to have more of LWdown gapfilled than rainfall. Again you can change the setting as per your requirements.

We already have a lot of the data processed so if you want the ready data, feel free to get in touch (a.ukkola@unsw.edu.au).

Cheers,
Anna