Nutrient loading and meteorological conditions explain interannual variability of hypoxia in Chesapeake Bay

We use geostatistical universal kriging and conditional realizations to provide the first quantitative estimates, with robust estimates of uncertainties, of the seasonal and interannual variability in hypoxic volume in Chesapeake Bay, covering early April to late October for 1985 to 2010, and explore factors controlling that variability. Results show that the time when the hypoxic volume reaches its maximum has moved from late to early July over the examined period, but that there is no trend in the seasonal-maximum hypoxic volume itself. No significant trend was found in the timing of onset of hypoxia, but the end of the hypoxic period has moved from October to September. Including nutrient loading from the Rappahannock River in addition to the Susquehanna and Potomac Rivers is found to be beneficial for explaining the interannual variability of hypoxia. Overall, January to May total nitrogen loads from these three rivers, April to August southwesterly and northeasterly winds, and April and May precipitation explain >85% of the seasonally averaged interannual variability in hypoxic volumes. Southwesterly winds affect hypoxia by increasing vertical stratification, while precipitation likely acts as a surrogate for nonpoint sources of nitrogen downstream from monitoring stations. The relative contribution of nutrient loading to the overall interannual variability suggests that 28-35% reductions in monitored nutrient loads may not be sufficient to achieve a corresponding reduction in hypoxic conditions as had been suggested in previous studies, at least in the short term.