You can use -

1. Del Vanoni (1975): (1) SDR =0.4724*A^(-0.125) where A= watershed area (km²)

2. Exposed model by USDA (1972): (2) SDR = 0.5656*A^(-0.11) where A= watershed area (km²)

3. Boyce model (1975): (3)  SDR = 0.3740*A^(-0.2382) where A=watershed area (km²)

You can follow this article

https://pdfs.semanticscholar.org/b5c4/db8a92e3ca8516704aa477e083fd0b675c4a.pdf

You might try Roehl (1962) based on sediment delivered to ponds, lakes by drainage size and physiographic area. I think the Water Erosion Prediction Project (WEPP) has a delivery approach. Some stream types are efficient delivery systems (such as gullies or high gradient channels), and some channel types are not, such as braided streams are inefficient at sediment delivery and therefore aggrade. If any model has been validated for your area of interest with erosion and sediment delivery data by hydrologists, that one should be considered. Headwater drainages with channel gradients over 5% have more sediment delivery efficiency due in part to limited floodplain, while low gradient valleys and channels typically have floodplain access and greater ability for sediment storage. In one cooperative project (USFS, Duke Power, Atkins contractor), concerning gullied land, using RUSLE for activities, Roehl for sediment delivery, Rosgen BANCS Analysis of gully channel erosion with all delivered to channel system, we estimated about 90% of the sediment delivered was from the gully channel and bank erosion. RUSLE does not assess channel or bank erosion. RUSLE equation includes factors that are sometimes difficult to assess, such as P factor for effectiveness of BMPs in reducing delivery of soil erosion. I have used RUSLE in many types of environmental analyses, to use as indicator or comparative tool, helpful to make informed decisions, and not suggesting accuracy and precision in numbers generated. The WEPP model was not fully developed for our use, but is a more demanding model for data inputs, but more capability to address both erosion and sediment delivery. If collecting detailed channel metrics and sediment data, the Rivermorph software used with Rosgen BANCS Model may be helpful In estimating if the channel at various stages can deliver its sediment load.

I uploaded to researchgate an EIS we prepared for a large restoration project (2013 Loblolly Pine Removal) in which we used RUSLE, Roehl, GIS data layers, existing and planned activities to compare treatment options and assess areas of excessive activities where adjustments were made in activity entry or other mitigation’s to reduce sediment and water yield increases for water quality, channel and fish habitat maintenance. In the methods and results, I tried to describe how RUSLE, etc. was used, assumptions and qualifications of numbers generated. Working with our GIS experts, we were able to develop a GIS process or model to assess estimates of erosion and sediment of a variety of land use and management activities, and use as a tool in disclosing estimated effects of management options or alternatives, and supplying information for informed decision making.

Dear Sumedh Rajkumar Kashiwar,

I attached a paper for calculate Sediment Delivery Ratio. You can use this paper.

The attached example analysis of arid area in Turkey attached and recommended by Dr. Koralay appears to be a very good example of the process, which sometimes involves making assumptions. The analysis in this instance suggested that a dam could fill in only ten years without conservation methods to reduce erosion and sediment. The values obtained and analysis are indicators to help make an informed decision, and not necessarily to be considered accurate, as major flood, severe drought denuding vegetation, or severe wildfire could occur and fill the dam much faster. Luna Leopold wrote a paper that the severest erosion was not during wet periods, but during arid periods when plant cover was depleted or overgrazed, leaving watershed barren. The example paper mentioned fixing gullies, a needed mitigation, and especially since the RUSLE does not address gully, rill or other types of channel erosion, or landslides, etc.

.

Dr @william f. Hansen, could you please upload the mentioned paper. Thank you

If using Roehl for SDR, the rate adjusts for physiographic area and by watershed size to point of delivery. The delivery rate typically reduces as drainage area increases. I think I still have copy of Luna Leopold paper, who was writing on SW USA climatic cycles of past, showing dry periods adding more erosion and sediment than wet periods, due to vegetation cover differences. I will attach when I get into computer and file storage. In addition you might consider, channel margins and floodplain also may develop more vegetation during wet periods, this adds to channel roughness, reducing velocity, more roots holding, more sedimentation and retention, less sediment delivery downstream.

Here are a couple papers of Leopold you may want. Plus Roehl.

William F. Hansen

Sir,

Thank you very much for sparing precious time to share your expert inputs and valuable answer.

Thanks & Regards

Necla Koralay

Sheikh Mohammad Famim Ahmed

Madam and Sir,

Thank you very much for your valuable answers and sharing research papers.

William F. Hansen regarding Roehl's equation: LogDR=4.50047 - 0.23043 log10 W - 0.51022 colog R/L - 2.78594 log BR. I'm a bit confused regarding his use of Log, Log10 and Ln. I thought that Log and Log10 was the same thing, but when I try the formula with log and log10 as the same thing, it does not make sense. Do you have to use the Log as and Ln fuction? And subsequently the Colog as an CoLn? When I replace the Log for Ln, it gives me a proportional value, which would make more sense in this regard.

Regards,

Maarten