Hello,
I am working with atmospheric nitrate data, and I found significant data points with negative values (not missing or below the detection limit). I was wondering how to deal with them without removing them.
Not an expert in atmospheric chemistry, but when having negative numbers, these are often representing the magnitude of a loss or deficiency in atmospheric science, thus getting nitrate data below the detection limit might mean that, for the specific region of study, you might have fairly small concentrations of nitrate chemical in the atmosphere. Removing the data would create falsification or might pose a bad manipulation of your results as you would be omitting useful information which might even be an enormous finding.
Best
I'd say that these negative values are linked to measurement noise and its propagation in the retrieval. Normally, one shouldn't remove them to avoid creating positive bias. Of course, the instantaneous negative value does not have any physical sense. But, if you average over sufficiently large area or period, the negatives should disappear and you'll get a reasonable physical value. The other way of dealing with these values is to average the signal prior to retrieval, but this might be not an option in your case.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication