The overall goal of this report was to understand river discharge variability to improve conventional water management practices of Mexico's northern subtropical rivers. This report addresses whether: a) river discharge tendencies, patterns and cycles can be detected with proxy and instrumental records; and b) annual discharge can be forecasted by stochastic models. Eleven gauging stations of six major rivers; three lowland rivers discharging into the Pacific Ocean (Rios Santa Cruz, Acaponeta, and San Pedro); five upland rivers draining into the Pacific Ocean (Rio San Pedro: Peña del Aguila, Refugio Salcido, San Felipe, Vicente Guerrero and Saltito), one river flowing across the interior Basin (Rio Nazas: Salomé Acosta) and two more rivers discharging into the Northern Gulf of Mexico (Rio San Juan: El Cuchillo and Rio Ramos: Pablillos) were statistically analyzed. Instrumental recorded daily discharge data (1940-1999) and reconstructed time series data (1860-1940) using dendrochronological analysis delivered annual discharge data to be modeled using autoregressive integrated moving average, ARIMA models. Spectral density analysis, autocorrelation functions and the standardized annual discharge data evaluated annual discharge frequency cycles. Results showed ARIMA models with two autoregressive and one moving average coefficient adequately project river discharge for all gauging stations with four of them showing significant declining patterns since 1860. ARIMA models in combination with autocorrelation and spectral density techniques as well as standardized departures, in agreement with present (2002-2010) observations, forecast a wet episode that may last between 9 and 12 years thereafter entering again into a dry episode. Three dry-wet spell cycles with different time scales (1-2 years; 4-7 years; 9-12 years) could be discerned from these analyses that are consistent for all three northern Mexico's river clusters that emerged from a multivariate analysis test.
Cycles; patterns; tendencies; spectral density; drought and wet episodes