I am doing a study on small dam which have catchment area of 152 km 2. There is only one meteorological station installed in catchment area. There are two other meteorological stations which are near the Dam site but not in Dam catchment. The elevation of Dam is 570 meters. The only one Meteorological station which is within catchment area have elevation 2170 meters. The other two Meteorological stations which are near dam have elevation 543 meters and 517 meters. As the catchment have elevation difference ranging between 570 meters to 2170 meters. How I can find the actual rainfall of catchment.
Muhammad
can you perhaps use Spatial Interpolation as per the description below
https://docs.qgis.org/2.2/en/docs/gentle_gis_introduction/spatial_analysis_interpolation.html
FC
Hi,
You are so lucky having the stations exactly in the lower and upper bound of your catchment. As far as I know, rainfall (and temperature) have a very good correlation with elevation, so according to the accuracy you need, you can try one of the ways below:
1- simply averaging two of the stations(one of the lower ones, and the only high one)
2- using thiessen polygons method with the three stations
3- using one of the interpolation methods in GIS (Guess due to the few number of stations, IDW works better than others)
4- exploring the precipitation-elevation relation in your area by using some further stations (or maybe you could find it in papers) and converting the value of cells of your DEM to rainfall and finally averaging the rainfall on catchment cells.
the last one is the most accurate, IF that relation is well explored.
Good luck
Dear Muhammad,
I definitely agree with Hamed Khorasani's answer. You can try to investigate the relationship between elevation and long-term precipitation totals by means of a regression analysis. Using DEM cells as independent variables and the regression equation, you can obtain the estimates of precipitation totals for each of those cells (using matrix algebra or whatever it is called in your GIS), from which you should be able to compute the precipitation totals related to the whole catchment (using the sum over the whole set of cells belonging to your catchment divided by the catchment area).
The above mentioned procedure can be performed not only for the long-term precipitation totals but also for each month separately (resulting in several regression equations) if you want to form a time series. It is only a question of programming and batch processing skills. Note this procedure is not recommended for daily time steps since convective events may bias your estimates to a large extent.
Good luck with your work!
Ondrej from Czechia
Firstly, you will never find the actual rainfall for the catchment. All you can do is get an estimate of the rainfall, which by definition will have uncertainty. The goal is to minimise the uncertainty. If you do not extend the available data, then given the small number of gauges, the best option is to use the available data to determine the variation in the long-term average rainfall with elevation (see Hamed's comment).
https://www.researchgate.net/publication/264006114_Water_resources_of_Cyprus_under_changing_climatic_conditions_Modelling_approach_validation_and_limitations
You can then use a DEM to estimate the areal rainfall. The fact that the two low elevation gauges are not in the catchment is not really a problem. Even if you had a large number of gauges, I would still recommend that you include gauges outside the catchment.
Given the small number of gauges, getting reliable results from interpolation methods is going to be difficult, and I would advise against this route.
You could see if there are additional datasets which could help you. This includes collecting data from areas even further from the catchment you are studying, though you will need to think about the spatial correlation distance.
Looking at much larger catchments in India, we used a hybrid approach linking satellite data with gauge data:
https://www.researchgate.net/publication/256464473_Evaluation_of_approaches_for_estimation_of_rainfall_and_the_unit_hydrograph
However, given the small area of your catchment, such an approach may not be that useful.
Article Water resources of Cyprus under changing climatic conditions...
Article Evaluation of approaches for estimation of rainfall and the ...
I mostly agree with the other colleagues. Only one thing, for relying on interpolation you need to check quality of data. Precipitation might be simple to measure, but I have found that it is not. Also I wonder if the 2170 m elevation site gets snow, in that case undersampling might be even more pronounced. Ancillary information such relative humidity and even hidrological data might help you to check if you are missing precipitation events.
Hi,
In addition to assess data quality (missing observations and inhomogeneities in time series) in the only 3 available stations you have, I think also you need to figure it out if their timestep and period of record is appropriate for the hydrological analysis you have in mind. Hopefully it would be nice if you consider additional sources of climate information coming from satellites and reanalysis.
I also suggest to use CHIRPS (US Geological service) which has a 0.05 resolution and covers a huge region of the planet, 50S - 50N. The monthly and daily series have more than 30 years (1981- up to date) and combine satellite data + gauge stations around the world. Visit here:
http://chg.geog.ucsb.edu/data/chirps/
Other product that may be of your interest is the NCEP Climate Forecast System Reanalysis (CFSR). See the following 3 links:
http://rda.ucar.edu/pub/cfsr.html
http://onlinelibrary.wiley.com/doi/10.1002/hyp.10073/abstract
http://nomads.ncdc.noaa.gov/docs/CFSR-BAMS-2010.pdf
I am much thankful for all researchers guidance. Can any one share a paper in which catchment rainfall data is obtained by averaging two of the stations??
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication