A nitrogen (N) mass‐balance budget was developed to assess the sources of N affecting increasing ground‐water nitrate concentrations in the 960‐km2 karstic Ichetucknee Springs basin. This budget included direct measurements of N species in rainfall, ground water, and spring waters, along with estimates of N loading from fertilizers, septic tanks, animal wastes, and the land application of treated municipal wastewater and residual solids. Based on a range of N leaching estimates, N loads to ground water ranged from 262,000 to 1.3 million kg/year; and were similar to N export from the basin in spring waters (266,000 kg/year) when 80‐90% N losses were assumed. Fertilizers applied to cropland, lawns, and pine stands contributed about 51% of the estimated total annual N load to ground water in the basin. Other sources contributed the following percentages of total N load to ground water: animal wastes, 27%; septic tanks, 12%; atmospheric deposition, 8%; and the land application of treated wastewater and biosolids, 2%. Due to below normal rainfall (97.3 cm) during the 12‐month rainfall collection period, N inputs from rainfall likely were about 30% lower than estimates for normal annual rainfall (136 cm). Low N‐isotope values for six spring waters (δ15N‐NO3 = 3.3 to 6.3‰) and elevated potassium concentrations in ground water and spring waters were consistent with the large N contribution from fertilizers. Given ground‐water residence times on the order of decades for spring waters, possible sinks for excess N inputs to the basin include N storage in the unsaturated zone and parts of the aquifer with relatively sluggish ground‐water movement and denitrification. A geographical‐based model of spatial loading from fertilizers indicated that areas most vulnerable to nitrate contamination were located in closed depressions containing sinkholes and other dissolution features in the southern half of the basin.

Spring Creek Springs and Wakulla Springs are large first magnitude springs that derive water from the Upper Floridan Aquifer. The submarine Spring Creek Springs are located in a marine estuary and Wakulla Springs are located 18 km inland. Wakulla Springs has had a consistent increase in flow from the 1930s to the present. This increase is probably due to the rising sea level, which puts additional pressure head on the submarine Spring Creek Springs, reducing its fresh water flow and increasing flows in Wakulla Springs. To improve understanding of the complex relations between these springs, flow and salinity data were collected from June 25, 2007 to June 30, 2010. The flow in Spring Creek Springs was most sensitive to rainfall and salt water intrusion, and the flow in Wakulla Springs was most sensitive to rainfall and the flow in Spring Creek Springs. Flows from the springs were found to be connected, and composed of three repeating phases in a karst spring flow cycle: Phase 1 occurred during low rainfall periods and was characterized by salt water backflow into the Spring Creek Springs caves. The higher density salt water blocked fresh water flow and resulted in a higher equivalent fresh water head in Spring Creek Springs than in Wakulla Springs. The blocked fresh water was diverted to Wakulla Springs, approximately doubling its flow. Phase 2 occurred when heavy rainfall resulted in temporarily high creek flows to nearby sinkholes that purged the salt water from the Spring Creek Springs caves. Phase 3 occurred after streams returned to base flow. The Spring Creek Springs caves retained a lower equivalent fresh water head than Wakulla Springs, causing them to flow large amounts of fresh water while Wakulla Springs flow was reduced by about half.

No abstract

Methods for estimating the magnitude of floods for selected percent chance exceedance probabilities are presented for ungaged streams in Florida that are not sub stantially affected by regulation, channelization, or urban development. Flood-frequency flows also are presented for 275 Florida streamgages used in the regional regression analysis. Regression relations used generalized least-squares regression techniques to estimate flood magnitude and frequency on ungaged streams as a function of basin drainage area and a storage factor. These regression equations were developed for four different hydrologic regions in Florida. The flood regions were delineated based on plotted residuals, previous floodfrequency studies, and geologic, physiographic, and drainagearea maps. The methods used in this report are based on floodfrequency characteristics for 305 streamgages including 275 in Florida and 30 in the adjacent states of Georgia and Alabama, all having at least 10 years of record through September 2006. For the larger streams outside the limits of the regression equations-the Apalachicola River and Suwannee River at Ellaville and below-the report includes graphical relations of peak flow to drainage area. Previous Studies This report supersedes previous USGS reports estimating the magnitude and frequency of flooding in Florida. Earlier USGS studies describing flood-frequency relations throughout the State include reports completed by Pride (1958), Barnes and Golden (1966), Rabon (1971), Seijo and others (1979), Bridges (1982), Giese and Franklin (1996), and Hammett and DelCharco (2005). Most of the general organization, content, equations, and explanatory material contained within this Florida report are consistent with the discussions and language used for three similar reports from Alabama (Hedgecock and Feaster, 2007), Georgia (Gotvald and others, 2009), and South Carolina (Feaster and others, 2009). Data and additional discussion unique to the magnitude and frequency of floods in Florida have been inserted where applicable with the original USGS report sources accredited or shown as "modified from." Pride (1958) conducted a study estimating and regionalizing floods in Florida using peak-flow data through 1953 based on calendar year from 146 streamgages. Flood-frequency relations for the 146 streamgages were defined by graphical means. The State was divided into two flood regions based on

No abstract