Title: Improving Curve Number Runoff Estimates Using Dual Hydrologic Soil

Classification and Potential Contributing Source Areas Delineation Methods

Title of Journal: Journal of Water Resource and Protection, 9, 20-39 published on 2017

Author(s): Katherine Miller, Katherine Folk Clancy

I. Introduction

Runoff curve number (Curve Number) is predicting direct runoff or infiltration from
rainfall excess by means of observations and gathered data. Excess runoffs are
produced by direct precipitation which causes shallow subsurface flow, base flow and
overland flow. Overland flow is generated by infiltration and saturation excess flow.
Infiltration excess flows occur when water abruptly enters a soil system that the soil can
absorb it. Saturation excess flows are due to the soil being saturated with water that it
cannot absorb more water. Since direct precipitation is assumed to be negligible
contribution to direct runoff, overland flow is the aspect focused on most runoff models.
Runoff curve number models are used to be able to control both saturation and
infiltration excess flows.

II. The Problem

Curve number depends on the land use and soil type within the area under study.
Runoff models such the curve number can be inaccurate due to the initial abstraction
value within the CN model and the watershed delineation. Another factor that affects the
CN value is the hydrologic soil type. Wetlands and karst topographies are other things to
consider in the complexity of modeling runoff.

III. Methodology

Two methods to account for the inaccuracy of runoff models have is

1. To adjust the initial abstraction value within the CN model.

2. To improve the watershed delineation in order to better account for internal
drainage. Delineation methods such as Standard Fill Method (SFM), Cut Method
(CM) and Potential Contributing Source Areas Method (PCSAM) will be used in
this study.

The data for this study are also gathered on eighteen different watersheds. These
watersheds located in karst topographies and with the presence of wetlands.
Watersheds were chosen based upon the following criteria: 1) ten years of USGS
gaging station daily flow observations; 2) ten years of corresponding daily precipitation
data from NOAA near the USGS gage; and 3) a range of land cover types.
IV. Data Analysis and Discussion

The data gathered through SFM and CM show similar land use. This may present
errors and the representation through SFM and CM does favor any particular land cover
type. Regarding the wetlands, the PCSAM and SFM vary significantly. This variation is
due to the areas PCSAM covers. PCSAM only gathers data areas near the stream with
direction connection to the drainage areas.

Overall, PCSAM is very useful in large storms. PCSAM had significantly lower error
than that of SFM and CM. The watersheds with higher percentage of forest, wetlands,
and agricultural land use have closer value between 50 to 80 percent when using
PCSAM curve number model.

V. Conclusion

It can be concluded that the SFM and CM do not differ greatly. CM is not an
advisable method in improving delineation. In areas with more than 30 percent forest,
wetland or agricultural land use, it is better to use PCSAM. Delineation process for
watershed is also better in large precipitation events using undrained hydrologic soil
classification.