Multivariable statistical regression models of the areal extent of hypoxia over the Texas-Louisiana continental shelf

Observations of the areal extent of seasonal hypoxia over the Texas-Louisiana continental shelf from 1985 to 2010 are correlated with a variety of physical and biogeochemical forcing mechanisms. Significant correlation is found between hypoxic area and both nitrogen load (r(2) = 0.24) and east-west wind speed (r(2) = 0.16). There is also a significant increasing trend in the areal extent of hypoxia in time; a linearly increasing trend over the entire record (r(2) = 0.17), a step increase in area for the years 1994 and beyond (r(2) = 0.21), and a step increase for 1993 and beyond (r(2) = 0.29) were all found to be significantly correlated with area. The year 1988, often included in other studies, was found to be a statistical outlier, in that the statistical regression properties are strongly modified when this year is included. The exclusion of any other year does not have as great an effect as excluding 1988 from the record. The year 1989 is also excluded, as this year had no full shelf survey, for a total of 24 years of data for the record. Multivariable regression models using all possible combinations of the forcing variables considered were calculated. The best performing models included east-west wind, either a linear trend in time or step in time (1994 and beyond), and either nitrogen load or river discharge combined with nitrogen concentration. The range of adjusted correlation coefficients ranged from r(2) = 0.47 to 0.67. The best model (east-west wind, a step increase in time 1994 and beyond, river discharge, and nitrogen concentration) has a standard error of 3008 km(2).