text stringlengths 0 6.44k |
|---|
the Florida Keys reef tract after the 1980’s; and sustained nutrient concentration decline |
in waters leaving the Everglades (Briceño and Boyer 2007, 2008). |
Recent changes in BB have been reported by Migliaccio et al. (2008) who |
studied biogeochemical trends from years 1996 to 2007, using DERM, NPS and FIU |
data. They reported increasing trend for chlorophyll-a likely responding to the increase |
in NH4-N, SRP, and TPO4-P, in turn responding to a dramatic population increase and |
land use change in the watershed from agricultural land to residential and urbanizing |
areas. |
Our analysis of time-series components was performed with Akritas Theil/Sen |
regression (ATS) for secular trend (Appendix 9) and Zcusum charts for detailed |
scrutiny. A Cusum chart is a plot of the cumulative sum of deviations (Sn= Σ[zi-T] for |
i=1….n) from a target (T), in our case, the time-series grand mean, against n, the |
sample number (or date if regularly sampled) (Ewan, 1963). Cusum charts and Cusum |
analysis are standard procedure in the field of industrial process control (Duncan 1974; |
Grant and Leavenworth 1980; Montgomery 2001). For a detailed explanation on |
51 |
construction, properties and applications of Zcusum charts, the reader is referred to |
Appendix 4). Visual inspection of Zcusum charts gives a detailed history of the |
parameters variability along the time-line or another variable gradient providing more |
information than ATS, ordinary linear regression or Line of Organic Correlation. |
Finally, we must always keep in mind that trend analysis is limited to the window |
of observation; trends change with continued data collection. In addition, water quality in |
South Florida may fluctuate according to disturbance events of longer periodicity, and |
trends may even reverse during a given period of record. |
Total Nitrogen |
As shown in Fig 3.19, TN declined across the Bay displaying two cycles common |
to all segments during the POC. From 1996 to 1999 values were in general above |
average (positive Zcusum slope), followed by a decline to below average values until |
2002, except in SCO where the decline began in 1998. The second cycle extends from |
2003 to 2007, with the break from above to below average approximately occurring at |
the end of 2004. In MBS the second cycle is less developed. Finally, seasonality is |
crudely developed, perhaps reflecting managed water deliveries non tuned to natural |
rain cycles and water inflow. |
Total Phosphorous |
Phosphorous displays a common pattern to all segments in Biscayne Bay, except MBS, |
consisting of (Fig 3.20): |
Below average TP values until the end of 1998, followed by above average TP to |
early-mid 2001 |
Seasonal changes with below average values until early 2002 followed by above |
average TP to late 2002 |
52 |
Figure 3.19: TN Zcusum charts for water quality segments in Biscayne Bay |
Below average TP until early 2005 followed by a long above average period, |
which started with a sudden increase in concentration in the summer 2005 |
(Wilma impact), followed by a less steep decline until mid-to-late 2006 when |
again a sudden increase in TP occurred. |
Finally, TP declines from early 2007 to the end of the POC. |
The extreme impact of hurricanes Katrina, Rita and especially Wilma in 2005, distorted |
the TP pattern at MBS (Fig 3.20) which otherwise would follow a similar pattern as the |
rest of the Bay. In this context, TP across all Biscayne Bay was declining until 2005. As |
for TN, seasonality is not well defined for TP either. |
-5 |
0 |
5 |
10 |
15 |
20 |
25 |
30 |
35 |
Jan-96 |
Jan-97 |
Jan-98 |
Jan-99 |
Jan-00 |
Jan-01 |
Jan-02 |
Jan-03 |
Jan-04 |
Jan-05 |
Jan-06 |
Jan-07 |
Jan-08 |
Jan-09 |
TN Zcusum |
MBS TN |
-10 |
-5 |
0 |
5 |
10 |
15 |
Jan-96 |
Jan-97 |
Jan-98 |
Jan-99 |
Jan-00 |
Jan-01 |
Jan-02 |
Jan-03 |
Jan-04 |
Jan-05 |
Jan-06 |
Jan-07 |
Jan-08 |
Jan-09 |
TN Zcusum |
CS TN |
-5 |
0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.