Hello
I'm trying to analyze trends of drought characteristics (such as drought duration, magnitude...etc) extracted based on defined threshold limits. The data set, thus, included years with blank values because of being a non-drought year or excluded value below threshold limits.
I'm unsure how to treat these years. I find it illogical to consider it as missing value since it isn't in actual sense. Filling the blank years with zero value my result into wrong Mann-Kendall statistics and value of Sen's slop. Actually, when the gap between two consecutive values is large and filled with zero, Sen's slop becomes zero while the MK test result is significant.
I'm just wondering how to handle this challenge. I would appreciate if any one has a suggestion.
Thank you so much
David
Hell,
See the attached MAKSENS excel file and its manual.
In the MK trend test missing values are allowed and also the Sen’s method can be computed through existing missing data.
Also, Take a look on the provided suggestion in the following discussion:
https://www.researchgate.net/post/The_applicability_of_Mann-Kendall_trend_test_for_unequally_spaced_data_in_time
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts