Submitted to: Southern Conservation Agricultural Systems Conference
Publication Type: Proceedings
Publication Acceptance Date: June 28, 2010
Publication Date: July 22, 2010
Citation: Endale, D.M., Schomberg, H.H., Fisher, D.S., Jenkins, M. 2010. Curve numbers from a long-term no-till crop field in the Georgia Piedmont. 32nd Annual Southern Conservation Agricultural Systems Conference, July 20-22, 2010, Jackson, Tennessee. CDROM. Interpretive Summary: Models can be useful tools to assess environmental impacts of natural phenomena or actions by man. However, models need to be tested with preferably long-term independent data to insure that they perform well under a wide range of expected environmental and management conditions. Researchers at the USDA-ARS J. Phil Campbell Senior Natural Resource Conservation Center near Watkinsville, GA, used 33 years of rainfall and runoff data collected from a 6.5-acre watershed (P1) that had been managed under a no-till system to test USDA’s Curve Number (CN) model developed in the 1950s to estimate the portion of the rainfall that leaves a parcel of land as runoff. In comparison to curve number values of 60 to 70 expected from standard tables for the conditions at P1, derived mean and median CN values were 36 and 31, respectively, implying that there was less runoff than would have been expected from standard CN-based estimates. The commonly used model was not well adjusted to the location soil managed under no-till for long periods. Consensus is developing from research across many other locations that this model needs adjustment to provide more realistic prediction of runoff. The data indicated that this historically eroded watershed under long-term no-till management behalves more like a well kept pasture or meadow than a cropped one. The need for adjustment is not surprising since the original model was developed from limited data in comparison to the data that have been gathered since its development. Such long-term field data are essential for improving models. The Curve Number model is popular throughout the world and these observations should be of interest a wide variety of users.
Technical Abstract: Since its inception in the 1950s, acceptance, use and adaptation of the Curve Number (CN) method for estimating direct runoff from a rainfall event has increased worldwide. Some inconsistencies, limitations and problems have been identified as a result. There have been calls for development of locally defined CN values to address concerns with regional and seasonal variations. Researchers at the USDA-ARS near Watkinsville, GA, derived CN values from 33 years of rainfall-runoff data gathered from a 6.7 ac instrumented catchment (P1) managed under no-till. Summer crops included soybean, sorghum, millet, cotton and corn, with barley, wheat, crimson clover and rye as cover crops. Derived mean and median CN values were 36 and 31, respectively, compared to 60 to 70 expected from standard tables for the conditions at P1, implying that there was less runoff than would have been expected from standard CN-based estimates. Such low CN values suggest that, under long-term no-till management, hydrologic performance of P1 has become similar to those of pastures or meadows in good hydrologic condition which have low runoff potential. In half of the 126 storm events evaluated, <1% of the rainfall was portioned to runoff. On average, only 6.5% of the rainfall was partitioned to runoff. The derived median value of the initial abstraction ratio (') was 0.04, compared to the standard value of 0.2, which is very close to 0.05, a value proposed by researchers to improve the CN method. Such long-term data from field operations are essential for improving models.