that allows adjustable flex by changing the window size) that suggests trends. These precipitation gauges are 

 distributed across the Rincon Project drainage suppljnng the delta and possess the better quahty data for that 

 area. Annual precipitation recording began shordy after 1900, except for the Corpus Christi record that began 

 in 1948. The figures show the prominent drought of the 1950s (and several years of the late 1940s). The 

 Lowess cixrves show short-term recent rising precipitation at three of the four gauges. The Corpus Christi 

 Lowess curve portrays an unrealistic initial pattern because of analysis effects from initial data from the 

 drought. The inclusion of the drought years to the CPC baseline largely caused the positive trend in 

 precipitation. 



Line plots (not given) of the Corpus Christi WSO AP, Sabinal, and Beeville 5 NE gauges produced respective 

 positive, negative, and positive slopes using the available annual precipitation for each gauge. The Corpus 

 Christi WSO AP data yields a negative slope if precipitation is restricted to 1965 or more recent toformation. 

 The Beeville 5 NE precipitation produces a negative slope if restricted to 1960 or more recent. 



Asquith et al (1997) conducted a study to determine the ctirrent mean freshwater inflows to bay systems that 

 included the Corpus Christi Bay. As part of their study, they examined temporal trends in inflows. They 

 studied trends in precipitation of several precipitation gauges. Asquith et al (1997) applied the Mann-Kendall 

 test (Helsel and Piirsch, 1992; Hollander and WoLfe, 1973) to time series of seasonal and annual rainfall from 

 the Corpus Christi WSO AP and Refugio 7 North (located approximately 30 miles east of the Beeville 5 NE 

 gauge) gauges. The Mann-Kendall test is a rank-based nonparametric test similar to the familiar Wilcoxon- 

 Mann-Whitoey procedures to test for monotone increasing character in data. Results of the statistical analysis 

 indicated P- values not significant at the 0.05 level in all seasonal and annual time series of 1968-93 data for 

 Refugio 7 North, and similarly for Corpus Christi WSO AP data excepting winter (January, February, March) 

 that produced an upward trend with a P-value of 0.05. The annual Corpus Christi data produced a 0.25 

 P-value. 



Figure 6 presents the mean annual precipitation computed separately for the four gauges, Sabinal, Corpus 

 Christi, Beeville 5 NE, and Cotulla for each time period of interest to the Rincon Project. This portrayal of 

 gauge mean precipitation shows the change across time periods and gauges. Because the Corpus Christi record 

 began in 1948, the first period duration was restricted to 10 years; other period lengths were 24 and 18 years. 

 The figure shows all gauges displaying the same general pattern over the three periods, least precipitation during 

 the drought period, greatest amount during the second period, and intermediate amounts during the third and 

 most recent period. 



Precipitation in the Nueces watershed is not monotonic increasing over the three study periods. However, 

 there could be patterns of rainfall amounts that could cause more inundations of the Nueces Delta than other 

 patterns with similar annual precipitation. Namely, the annual precipitation could occur in relatively few large 

 amoimt events. This could occur from increased tropical disturbances numbers, more persistent frontal 

 patterns that favor precipitation in the watershed, or perturbations in a persistent moist Pacific flow across 

 northern Mexico and the southern United States (shown in Figure 9). 



Figures 7 and 8 present cumulative adjusted annual precipitations over the gauge record periods for the Sabinal 

 and Corpus Christi WSO AP gauges (through 1998), respectively. Annual precipitation values were adjusted by 

 removing the long-term annual mean. Difference values were then summed in time and plotted. The 

 cumulative precipitation differences are related to soil moisture. Though such cumulative difference figures are 

 impacted by the choice of data starting point, the figures for the Rincon Project study area portray the 

 substantial annual precipitation deficit created by the severe drought of the late 1940s and 1950s. The longer- 

 term Sabinal record shows the contrasting conditions prior to and after the 1950s. The plot for Sabinal 

 suggests that the upper Nueces River basin may have frequendy been in below-normal soil moisture conditions. 



ENSO AND TROPICAL CYCLONES IN THE ATLANTIC BASIN 



The relative maximum in precipitation that occurs in the fall season in the Nueces River watershed is largely the 

 result of, (1) mid-latitude perturbations that in fall strengthen, adequate to reach south Texas as the subtropical 

 and polar jets reestablish in the stronger north/south temperature gradient, and (2) tropical perturbations that 

 impact the western Gulf of Mexico. Tropical cyclones can substantially impact western Gulf precipitation. 

 Landsea et al. (1999) studied Atlantic hurricanes. United States (lower 48 States) land-falling hurricanes and U.S. 

 normalized damage (adjusted for changes in inflation, coastal county population, and wealth) time series for 



Appendix D ♦ D-3 



