I have tried searching and find that the articles and books I came across do not give a clear explanation on determining the level of streamflow variation based on CV.
I am trying to distinguish the variation level of my mean annual flow data series, which value ranges from 0.31 to 1.49 (+ one outlier: 3.59). I was just going to consider 1.0 is high, but I found a paper that said the general rule use 1.0 as the high level, and another source that said
I am not prepared to give you references. CV is a standard statistical calculation. My concern relates to outliers, which are often important in long term studies relative to characterizing droughts and floods, extremities of wet or dry years, but obviously will not plot well in short term data sets or if one desires estimates within the normal range, disregarding extremes. Whether the data set is 20 years or 100 years, it is possible to have a embedded a 500 or 1000 year event, that seems to throw standard curves off from what we want to see. We must realize that our projections may suggest the outlier return period, we are really only making an educated guess. And I remember wet year 2004 when the nearest stream gauge to our bank stabilization project recorded 8 events with bankfull and larger flow.
In my estimation, it is very important to disclose the presence of what may be outliers relative to the data duration, and how these outliers were included or excluded. Perhaps presenting results including and excluding outliers might be useful to presenting results or to others. Discarding outliers In data sets needs to be disclosed. Dr. Luna Leopold identified in one of his papers about SW USA, that droughts can be cyclic, lasting about 20 years. Amazingly, due to the loss in ground cover, these periods can have the highest erosion.
Hydrology studies typically use water years, with separation of each year in the dry season, such as October 1 to September 31. It is possible in some years to have an overload of rainfall near the end of September, resulting in an abnormal soil moisture storage in one year, that discharges into the next year. So sometimes separating water year dates can separate the severe rainfall event at least partially from its discharge event. For some types of studies, the outlier becomes an important part of the results. But it is not unusual for variable conditions in many areas.
Well,
Have a look to how using Cv% in the phenomena interpretation as found in the attached article.
Article Hydrochemical model and probable pollution in the water of E...
Regards
I am not prepared to give you references. CV is a standard statistical calculation. My concern relates to outliers, which are often important in long term studies relative to characterizing droughts and floods, extremities of wet or dry years, but obviously will not plot well in short term data sets or if one desires estimates within the normal range, disregarding extremes. Whether the data set is 20 years or 100 years, it is possible to have a embedded a 500 or 1000 year event, that seems to throw standard curves off from what we want to see. We must realize that our projections may suggest the outlier return period, we are really only making an educated guess. And I remember wet year 2004 when the nearest stream gauge to our bank stabilization project recorded 8 events with bankfull and larger flow.
In my estimation, it is very important to disclose the presence of what may be outliers relative to the data duration, and how these outliers were included or excluded. Perhaps presenting results including and excluding outliers might be useful to presenting results or to others. Discarding outliers In data sets needs to be disclosed. Dr. Luna Leopold identified in one of his papers about SW USA, that droughts can be cyclic, lasting about 20 years. Amazingly, due to the loss in ground cover, these periods can have the highest erosion.
Hydrology studies typically use water years, with separation of each year in the dry season, such as October 1 to September 31. It is possible in some years to have an overload of rainfall near the end of September, resulting in an abnormal soil moisture storage in one year, that discharges into the next year. So sometimes separating water year dates can separate the severe rainfall event at least partially from its discharge event. For some types of studies, the outlier becomes an important part of the results. But it is not unusual for variable conditions in many areas.
Thank you for the valuable responses William F. Hansen , Ankur Srivastava , Bayan Hussien.
I will study your suggestions.
Dear Adam,
High Annual Variable River (Cv > 0.35); Low Annual Variable River (Cv < 0.35)
The reference is:
Adeloye A J 2012 Encyclopedia of Lakes and Reservoirs, ed L Bengtsson, et al. (Dordrecht: Springer Netherlands) pp 346-55
Kind regards,
Issa
Another idea if you want to visually compare stream gauging stations as to their responses, such as across a region would be to normalize the flow data by area (such as changing cfs to CSM, cubic feet per second per square mile). Then take the normalized flow data for the various stations and plot in flow duration curves (can be easily done in MS excel). When I was in Georgia, I plotted all the stream gauging stations within or near the National Forest, and this displayed the differences and similarities among the Ridge and Valley, Blue Ridge Mtns and Piedmont stations. The flow duration curves when plotted easily show how stations compare for mean flow, low flows, and moderately high flows. Being able to display and visualize these differences may help in explanation When discussing variability.
- Badges
- Science topic
- Similar topics
- Geoscience
- Hydrology