Curve number estimates for conventional and conservation tillages in the southeastern Coastal Plain

The USDA Natural Resource Conservation Service curve number (CN) method for estimating surface runoff is frequently used in natural resource modeling. Water yield and subsequently water quality estimates depend heavily on CN selection. This study was conducted to estimate CNs for a cotton-peanut rotation under conventional and strip tillage (ST) methods for growing and dormant seasons. A comparison between alternative methods for calculating CN and their applicability was also made. Rainfall-runoff data measured from 1999 to 2005 at a field study site in South Georgia were used to calculate CNs by averaging, lognormal, and data-censoring methods. For conventional and STs, CNs by the averaging method using year-round data were 89 and 84, respectively, and by the lognormal method were 89 and 83, respectively. Results from the data-censoring method were 81 and 75, respectively, which matched standard table values developed from a long-term series of annual maximum runoff. Values were also found to vary by season. Curve numbers by the lognormal method for ST were 83 and 88 for growing and dormant seasons, respectively; however, there was no difference between growing and dormant seasons, 89, for conventional tillage. The corresponding CNs by the data-censoring method for ST were 71 and 79 for growing and dormant seasons, respectively, and for conventional tillage were 82 and 79 for growing and dormant seasons, respectively. Based upon errors of the estimates, runoff estimates showed no improvement when separate CNs for the two seasons were used. The data-censoring method CNs yielded lower runoff estimate errors than CNs obtained by the lognormal method. The data-censoring method is recommended for determining CNs from plot rainfall-runoff data pairs.