1. 1 Quality Control of Phycoerythrin Data from The Columbia River Estuary Development of a Correction for Turbidity Artifacts. Observation ● Prediction ● Analysis ● Collaboration www.stccmop.org Sarah Riseman Acknowldegements: Antonio Baptista Charles Seaton Needoba/Peterson Lab Lydie Herfort CMOP Field Team CMOP Staff

2. 2 Background: phycoerythrin Chemotaxonomic pigment –Synechococcus spp. –Cryptophytes –Myrionecta rubra** Measured using Fluoromtry –Autofluorescent: fluorometers designed with filter sets to specifically measure the presence of phycoerythrin –Turbidity & temperature are two sources of bias of in situ fluorometric measurements

3. 3 Background: Basics of a Filter Fluorometer Adapted from:http://www.turnerdesigns.com/t2/doc/appnotes/998_0050/0050_c3.html

4. 4 Chlorophyll & Phycoerythrin Spectral Properties Chlorophyll a Phycoerythrin 460 523 620 - 715 578

5. 5 SATURN-03 Data from 2m (July 2010 – March 2011) Turbidity Chlorophyll Raw Phycoerythrin

6. 6 Combined Turbidity Correction Turbidity Chlorophyll Corrected Phycoerythrin

7. 7 Preliminary Turbidity Correction AugustJulySeptember *NovemberOctober Monthly Phycoerythrin vs. Turbidity, Colored by Chlorophyll Saturn03 - 2m Turbidity Phycoerythrin

8. 8 Preliminary Turbidity Correction combined remainder of data into single data set time period not used in regression (8/15 – 10/10) Time

9. 9 Preliminary Turbidity Correction – Linear Fits Low (< 225ntu) high(225-500ntu) All Data

10. 10 Preliminary Turbidity Correction

11. 11 Preliminary Turbidity Correction back Turbidity Corrected Phycoerythrin Chlorophyll

12. 12 Combined Turbidity Correction Turbidity Chlorophyll Raw Phycoerythrin

13. 13 New Turbidity Correction

14. 14 Details of Correction Determination 1.Data Selection: - winter only - all 3 depths - chlorophyll < 2 ug/L -turbidity < 625 NTU 2. Data Binned & Averaged -1 NTU turbidity bins -data within mean +/- 3 s.d. averaged 3. Curve Fitted -several evaluated -simplest selected: -Exponential rise to maximum: Single 3 parameter Exponent

15. 15 New Turbidity Correction Corrected Phycoerythrin Turbidity Chlorophyll

16. 16 Comparison of Turbidity Corrections

17. 17 Conclusions The turbidity corrections need to be periodically evaluated for changes and updated accordingly With correction, the phycoerythrin data display a clear pattern corresponding to Chlorophyll These data clearly identified the onset of the 2010 M. rubra bloom With field samples a calibration to cell abundance should be possible

18. 18 Discussion Topics: where do we go from here? Are the corrected data useful? - Can they be used to monitor the onset & development of M. rubra blooms in the CRE? How do we validate these signals? - field sampling sampling - monitoring sensor drift

19. 19 EXTRA SLIDES FOLLOW

20. 20

21. 21 Phycoerythrin/Turbidity Relationship Over Time RE-PLAY JULYAUGUSTSEPTEMBEROCTOBERNOVEMBERDECEMBERJANUARYFEBRUARYMARCH JULY 2010 – MARCH 2011

22. 22 Processing of SATURN-03 Data

23. 23 Preliminary Correction

24. 24 Turbidit y Chlorophy ll Phycoerythrin - raw

25. 25 Turbidit y Chlorophy ll Phycoerythrin - corrected

26. 26 Phycoerythrin - corrected Salinity

27. 27 Saturn01 Turbidity Correction based compiled data from 3 depths (2m,8m,13m) where chl < 1.5 ug/L during Aug 4 – Aug 22, 2010) Saturn01 Turbidity Factor Based on Bin/Avg Regression : TF = 1178.4*turb + 1334.1 Corrected PE = PE – TF (+2850 to raise to raw data baseline)

28. 28 Sat01 -2m PE Correction