D R A F T - 09/09/2014

Exploratory Analysis of Everglades Flow & Phosphorus Dynamics

Library of Time Series & Diagnostic Charts Linked to Google Earth

prepared for

U.S. Department of the Interior & U.S. Environmental Protection Agency

Introduction

The extensive long-term hydrologic and water quality data stored in DBHYDRO are valuable resources to support Everglades research and management to achieve water quality and hydrologic restoration goals. Interpretation of water quality data to support restoration management decisions and optimization of STA performance requires consideration of both concentration levels and their covariance with hydrologic conditions. For example, an apparent long-term trend in concentration could reflect a trend in flow or water level in response to climatologic variations, as opposed to a response to management measures or basis to forecast future conditions. A better understanding of the flow/concentration dynamics of the STAs is needed to develop operational plans to improve STA peformance. Reports and Powerpoints often contain two-dimensional time series charts or tables of water quality data that do not convey the underlying variability and its assocation with hydrologic conditions (flow, depth, rainfall) or other factors, such as season, changes in sampling methods, and changes in water management. Useful information in the data can be obscured by over-simplifications, decorations, and words.

An integrated database of flow and water quality data has been developed to support exploratory analyses & modeling of phosphorus dynamics in the Everglades Basin & Stormwater Treatment Areas. The platform allows rapid display of the data to reflect the covariance between hydrology and water quality while considering other dimensions of the data.

Variables include flow [ Q ], TP concentration [ C ], and TP load [ Q x C ]. Another metric, "Excess Load" [ Q x ( C - Cb ), >=0 ], reflects the potential for a given discharge to have adverse water quality impacts downstream, as measured by an increase in concentration above background or target levels (Cb). Other water quality variables, water depth, and stage and will be considered in future updates.

Spatial dimensions are associated with site location, water body, station type (structure vs. marsh vs. lake), and flow path in the STAs. Most of the longterm sites with flow and concentration data in and south of Lake Okeechobee are included.

Relevant time scales include daily, monthly, seasonal, yearly, cumulative, and long-term.

Correlations among flow, load, and concentration are relevant to development of regional water management strategies to improve hydrology without adversely impacting water quality. Similar correlations are relevant to the design and operation of the STAs to optimize performance.

Variations in sampling frequencies & methods (grab vs. composite) impact precision, accuracy, and appropriate computation procedures for loads and flow-weighted-mean concentrations derived from flow and concentration data. Consideration of these factors is especially important for evaluating performance, developing model input datasets, and optimizing monitoring networks for the STAs.

Methods for computing annual concentrations (flow-weighted vs. geometric) and water year definitions (May-April, Oct-Sept) impact compliance assessments.

Software previously developed to support construction of DMSTA calibration datasets (Walker & Kadlec, 2005; wwwalker.net/dmsta) has been adapted for this application. The software links and integrates flow and water quality data pulled from DBHYDRO to produce daily, monthly, and yearly time series of flow, flow-weighted and geometric-mean concentrations, and load for each monitoring site. Output datasets will be used to construct water and mass balances on the Everglades basins and develop input datasets to support DMSTA calibration. A library of charts that display the data along various dimensions has been generated as a general reference to support exploratory analyses and QA/QC of the paired flow and concentration data. To facilitate automation and expand graphics capabilities, the software platform (Excel / Access ) will be eventually be converted to the R programming language using templates recently developed for analyzing phosphorus trends in WCA-3A and ENP (wwwalker.net/ever_toc).

The Period of Record in the current database is May 2002 - July 2014. This reflects the period after refinements to SFWMD analytical procedures that lowered detection limits from 4 to 2 ppb and improvements to marsh sampling procedures. Flow and/or concentration are provisional and/or incomplete for many sites in 2014.

Chart Libraries

Three methods for accessing the output chart library are described below: Google Earth layers, PDF booklets, and Dashboards. These generally reflect an increasing level of detail. Linkage to Google Earth facilitates access and interpretation in a spatial framework that is ideal for analysis of Everglades datasets.

Everglades Basin:

Google Earth

PDF Booklet

Dashboard

WCA & ENP pump, structure, & marsh sites. Lake Okeechobee inflow, lake, & outflow sites.

Stormwater Treatment Areas:

Google Earth

PDF Booklet

Dashboard

STAs 1E, 1W, 2, 34, & 56. Permit inflows & outflows. Flow-way start, interior, end, seepage , & bypass sites.

The extensive library of charts accessed via Google Earth, bookets, or dashboards is designed to expedite exploratory analysis in a spatial framework. The intent is to develop a browser that will allow users to query subsets of the chart database for testing specific hypotheses or comparing sites in specific categories (e.g., trends in P load to all STAs and source watersheds or P dynamics in STA emergent vs. SAV cells.).

STA datasets can also be accessed via another portal (wwwalker.net/ever/stas) that supports modeling and data mining efforts under the Restoration Strategies Science Plan. The charts illustrate responses of outflow concentration and load to inflow pulses, which are considered to be a major factor limiting P removal performance and a focus of the Restoration Strategies projects and research.

Google Earth Layers

Everglades Basin

Stormwater Treatment Areas

Change map perspectives using Google Earth controls (mouse scroll -> zoom, shift-schroll -> tilt, ctrl-scroll -> rotate). Icon sizes are scaled relative to the long-term flow-weighted-mean TP concentration at each site. Icon colors reflect site categories (marsh, lake, structure, STA inflow, interior, outflow, seepage, etc...). Click on an icon to see a chart of monthly flow and flow-weighted-mean concentration at the corresponding site and links to more detailed information (Example for S8 below).

Site Popup:

Chart Grid (Other WCA-3A Inflows and Outflows):

Hyperlinks to more detailed information are listed below the popup chart. The links are also listed in the GE layer index on the left.

Daily

Daily time series of flow, load, and concentration. grab vs. composite samples.

Monthly

Monthly time series of flow, load, and concentration. seasonal variations.

Yearly

Yearly time series of flow, load, and concentration. data inventories by year.

Model

Regression model for predicting TP concentration as a function of flow, season, and trend.

Diagnostics

Depicting relationships among flow, load, concentration, & sample type on each time scale.

Trend - TP

Seasonal Kendall trend analysis - Grab TP Conc (WCA-3A & ENP sites)

http://www.wwwalker.net/ever_toc

Trend - Flow

Seasonal Kendall trend analysis - Flow (WCA-3A & ENP sites only)

http://www.wwwalker.net/ever_toc

DBHYDRO

Link to water quality station information from DBHYDRO.

All Sites

Grid of monthly time series charts for all sites in selected region (pdf file).

Facilitates comparison of TP and flow dynamics across sites in the same region.

Regions are defined by WCA (structures vs. marsh), ENP (structures vs. marsh), Lake, and individual STAs.

Use browser zoom controls or shortcuts (Ctrl +/-) to get the best view of the grid.

Click on any chart in the grid to get a pdf file with more detailed time series for the corresponding site.

The linked pdf files should appear in the Google Earth browser window after a brief delay as the file is downloaded. If the file does not load and Google Chrome is installed on your system, there is an option to load the file in Chrome (upper right corner). If Chrome is not installed, the same files can be accessed via the PDF booklets or Dashboards described below. Whether or not the file loads appears to depend on Google Earth constraints, PC hardware, and/or internet connection speed. This limitation is being further explored.

PDF Booklets

Everglades Basin

Stormwater Treatment Areas

PDF files with chart grids for each Everglades region or STA. Monthly time series of FWM concentration & flow (illustrated above). Scroll thru the datasets in your browser. Download the file and open in Acrobat to access bookmarks linked to each dataset. Click on any chart in the grid to get a pdf file with more detailed time series for the corresponding site.

Dashboards

Everglades Basin

Stormwater Treatment Areas

Dashboards provide links to chart grids for each Everglades sub-basin or STA in several formats. Columns contain data for different regions, STAs, and/or years. Each button links to file libraries or charts for a given dataset (columns) and theme (rows). Rows are color-coded to reflect different time scales or themes. An example for STA datasets is shown below as an illustration.

Row 1

Google Earth File

Rows 2-8

Library of chart pdf files by site.

Illustration of Chart Themes for S8

Rows 9+

Grids of charts for all sites. Click on any chart to download detailed time series for corresponding site.

Use browser zoom controls or shortcuts (Ctrl +) to expand dashboard panels or chart grids.

Google Earth File

Library of Chart Files for Each Site

Chart Grids - Data Inventories

Chart Grids - Daily Time Series

Chart Grids - Monthly Time Series

Chart Grids - Yearly Time Series

Chart Grids - Seasonal Variations

Chart Grids - Cumulative Time Series

Chart Grids - Percentage Changes vs. Time Frame

Details

This is work in progress. Results are undergoing additional QA/QC,especially for the STA sites.

DBHYDRO daily flow data are from MOD1 or Preferred DBKEYS when available, otherwise source DBKEYS are used.

DBHYDRO

Load units are kg unless otherwise noted. Concentrations are in ppb. Flow units are as labeled (cfs, kac-ft, or hm3).

Loads are computed by interpolating grab sample concentrations between sampling dates with positive flow. Regression models and other methods for computing loads based upon grab samples are also being explored. Both composite and grab samples are used to compute loads at sites with auto-samplers. Comparisons between grab and auto-sampler results are shown for different time scales.

WCA outflows (or ENP inflows) are used as surrogates for flow at marsh sites, where flow is not directly measured. Results depict temporal variations in potential P transport downstream, even though the load magnitudes are not accurate.The flow reference is appended to the marsh water quality site label. Integration of water depth data is planned. To the extent that flow is controlled by hydraulic resistance of vegetation, as opposed to structure operation, outflow volume would be a reasonable surrogate for average water depth.

TP loads are expressed as totals [ Q x C ] and excess loads above background [ Q x ( C - Cb), >=0 ]. Cb represents an assumed background concentration for everglades sites (6 ppb) or treatment target for STAs (13 ppb). When C is below Cb, the associated flow and total load would not have an adverse water quality impact downstream by increasing marsh concentration above background levels or increasing STA outflow concentrations above treatment goals. This could be an important metric for evaluating water quality impacts of hydrologic restoration projects that would add more flow to the system. Simulations using the Everglades Phosphorus Gradient Model (EPGM, http://www.wwwalker.net/epgm ) have indicated that the long-term average excess TP load (with Cb = 10 ppb) discharged into a marsh is approximately proportional to the area of impacted marsh area downstream, defined based upon criteria for water column P concentration, soil P concentration, and cattail density. Alternative values for Cb can be applied; for example, 8 ppb is the long-term goal for inflows into ENP's Shark River Slough and 100 ppb could be assumed as a target for STA inflows.

The software is structured to generate datasets and Google Earth files for other water quality variables, station subsets, time frames, and/or chart formats without additional coding.

Please send any questions or suggestions.

bill@wwwalker.net

http://www.wwwalker.net/ever/earth/index.htm

9/9/2014