Statistical Modeling of Spring Discharge at Rock and Wekiva Springs in Orange County, Florida. Special Publication SJ2007-SP12

摘要

The objective of this study is to develop a historic daily spring discharge time series for Rock and Wekiva springs from an assortment of auxiliary data such as: (a) previously recorded spring discharge at the spring of interest and at adjacent springs, (b) groundwater level measurements from adjacent monitoring wells, (c) lake levels from nearby lake-level gages and (c) rainfall data from nearby gauging stations. The study will investigate the correlation structure between various data types and test the applicability of simple multivariate linear models to generate daily discharge records based on these other variables for the common period of record. This report presents the results of data screening and preliminary statistical analysis for rainfall, groundwater level, and spring discharge data for Rock and Wekiva springs. It also explores the use of empirical models to provide estimates of daily discharge at these springs. These statistical models will take advantage of all available data to try to provide the most accurate estimates. In general, early time records are sparse and often not available for a number of locations. This will require the use of different models ranging in sophistication from simple correlation based models to multivariate regression models which can only be constructed when enough supporting data (e.g., rainfall and groundwater levels) are available at a sufficient number of nearby locations. These models will be used to run a continuous simulation model covering the period of record referenced by the constituent data. From the results of statistical modeling, standard flow-duration analysis for the system (discharge versus percent exceedance for the long-term simulation) will be conducted and standard high- and low flow frequency analyses for the system (frequency of spring discharge for set durations) will be carried out.