The network of precipitation and discharge measuring stations in the catchment upstream and near the project needs to be considered to assess the capacity of the same to adequately sample both spatially and temporally the precipitation and the stream flows. 

The measurement procedures and gap filling procedures in respect of missing data as also any rating tables or curves need to be critically examined so that they are  according to guidelines of World Meteorological Organization (WMO). 

Long-term data has to be checked for internal consistency between rainfall and discharges, as also between data sets by double mass analysis to highlight any changes in the test data for detection of any long term trends as also for stationarity. It is only after such testing that the data should be used for generating the long term inflows of water (volumes in 10 days, 15 days, monthly or yearly inflow series) into the reservoir. 

Sufficiently long term precipitation and run-off records are required for preparing the water inflow series. For working out the catchment average sediment yield, long-term data of silt measurement records from existing reservoirs are essential. These are pre- requisites for fixing the storage capacity of reservoirs. 

If long term run-off records are not available, concurrent rainfall and run-off data may be used to convert long term rainfall data (which is generally available in many cases) into long-term runoff series adopting appropriate statistical/conceptual models. In some cases regression analysis may also be resorted to for data extension. 

Estimation of Average Sediment Yield from the catchment area above the reservoir

It is usually attempted using river sediment observation data or more commonly from the experience of sedimentation of existing reservoirs with similar characteristics. 

Where observations of stage/flow data is available for only short periods, these have to be suitably extended with the help of longer data on rainfall to estimate as far as possible sampling errors due to scanty records. Sediment discharge rating curve may also be prepared from hydraulic considerations using any of the standard sediment load formulae, such as, Modified Einstein‟s procedure, Young‟s stream power, etc. It is also necessary to account for the bed load which may not have been measured. 

Bed load measurement is preferable and when it is not possible, it is often estimated as a percentage generally ranging from 5 to 20 percent of the suspended sediment load. However, actual measurement of bed load needs to be undertaken particularly in cases where high bed loads are anticipated. To assess the volume of sediment that would deposit in the reservoir, it is further necessary to make estimates of average trap efficiency of the reservoir and the likely unit weight of sediment deposits, along with time average over the period selected. 

The trap efficiency would depend on the capacity inflow ratio but would also vary with the locations of controlling outlets and reservoir operating procedures. Computations of reservoir trap efficiency may be made using the trap efficiency curves such as those developed by Brune and by Churchill (see IS: 12182-1987).