Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Understanding Salinity Variability in the Columbia River Estuary Sierra & Julia Observation ● Prediction ● Analysis ● Collaboration Frontline Mentor:

Similar presentations


Presentation on theme: "1 Understanding Salinity Variability in the Columbia River Estuary Sierra & Julia Observation ● Prediction ● Analysis ● Collaboration Frontline Mentor:"— Presentation transcript:

1 1 Understanding Salinity Variability in the Columbia River Estuary Sierra & Julia Observation ● Prediction ● Analysis ● Collaboration Frontline Mentor: Pat Welle Senior Mentor: Dr. Antonio Baptista

2 2 Center for Coastal Margin Observation and Prediction Collaboration of scientists aiming to improve the understanding of the Columbia River Estuary and Coastal Margins on a molecular and systematic scale National Science Foundation Center Partnership with OHSU, University of Washington, Oregon State University

3 3 Columbia River Estuary Border of Oregon and Washington Columbia River spills into the Pacific Ocean

4 4 Columbia River Estuary Second largest estuary in United States Columbia River flowing into the Pacific Ocean Transition zone Mixing between fresh and salt water Influence of tides 70% of fresh water from the Columbia River goes through Bonneville Dam

5 5 Saturn Observation Network Science and Technology University Research Network Combination of endurance stations and mobile sensors –Stations, drifters, gliders Includes numerical representation of Columbia River –DB11, DB14, DB16, DB22 Stations and models encompass estuary, plume and shelf

6 6 Model Set of mathematical equations that represent physical processes and properties applied over a chosen space. The space is broken down into multiple segments that form a grid. Salinity values are determined for each piece of the grid

7 7 Station Map Washington Oregon Pacific Ocean Sandi Am169 Cbnc03

8 8 Lower Sand Island light (sandi) Endurance Station Saturn Observation Network CT at 7.9 meters Salinity and temperature

9 9 Astoria-Megler Bridge South Channel (am169) Endurance Station Saturn Observation Network CT at 14.3 meters Salinity and temperature

10 10 Cathlamet Bay North Channel (cbnc03) Endurance Station Saturn Observation Network CT at 6.5 meters Salinity and temperature

11 11 Our Project Comparing simulated data versus observed data to understand salinity variability in the Columbia River Estuary and what causes the differences between what the model predicts and what the data shows. Salinity (psu) April 30 th - May 13 th 2009 AM169 Week 18-19

12 12 Forces in the Estuary Tides –Mixing of salt and fresh water and also effects the salt water intrusion upstream of the mouth River discharge –Salt water intrusion Wind –Upwelling and Downwelling

13 13 Tides Tide Cycle: –12.4 hours between high and low tide Spring tides –Occur during full and new moons –Low salt water intrusion Neap Tides –Occur during quarter moons –High salt water intrusion Salinity (psu) April 16 th - April 20 th 2009 Week Tides

14 14 River Discharge Majority of fresh water in the estuary flows through Bonneville Dam, 140 miles east of estuary Fresh water not flowing through Bonneville, comes from Willamette River, other forms of precipitation, tributaries

15 15 Coastal Upwelling Wind blows from north along the coast in a southern direction Usually occurs during summer months Upwelling causes more salt water intrusion during summer months Surface Water Movement

16 16 Wind Surface water sinks Coastal Downwelling Wind blows from south along the coast in a northern direction Usually occurs during winter months Downwelling causes less salt water intrusion during the winter months North South

17 17 Procedure: MATLAB MATLAB –Data analysis tool, similar to Excel –Graphing –Statistical analysis –Commands –Workspaces 1.Import data from database into MATLAB using pgAdmin or PuTTY 2.Remove bad data(clear NaNs) 3.Interpolate the observed to the model data 4.Graph data

18 18 PuTTY & pgAdmin Programs to access data from database through systems of queries and commands Data is imported into MATLAB for use pgAdmin PuTTY

19 19 MATLAB

20 20 MATLAB

21 21 Smoothing Data Takes data points and uses a moving average function to smooth them over a specified period of time –Usually over a day or week Salinity (psu) July 16 th - August 13 th 2009 Smoothed Data

22 22 Time Series Project Creating plot configurations which include: –A comparison between modeled and observed salinity at stations Sand Island, Astoria-Megler Bridge, and Cathlamet Bay –Discharge –Tides –Wind velocity From west to east 2 weeks 4 weeks Annual = Stations we focused on

23 23 2 Weeks Objective: To view short term trends between tides, discharge, wind direction and salinity values Graphs of sandi, am169, cbnc03 Graphs of tides, discharge and wind velocity

24 24 2 week page sandi discharge wind Tide cbnc3 am169

25 25 2 week: Conclusions Sandi SandI –Minimum simulated values are less than observed values by 2-5psu –Maximum simulated values 0-3psu less than observed values –Most accurate of the three stations Salinity (psu) September 3 rd - September 16 th 2009 Week sandi

26 26 2 week Conclusions am169 Am169 –Simulated values show similar trends as observed values but incorrect values –Simulated values are more accurate during the transitions from spring to neap tides, and are less accurate during transitions from neap to spring tides Pattern nonexistent during periods of high discharge –Simulated values are most accurate during periods of low discharge with spring tides SpringNeap Salinity (psu) September 3 rd - September 16 th 2009 Week am169

27 27 2 week: Conclusions cbnc03 Cbnc03 –Salinity values are greater during the transition from neap to spring tides and decrease during the transition from spring to neap tides Occurs only during low river discharge –February 5 th - End of March; simulated values indicate increased salinity when observed values indicate little or no salinity SpringNeap Salinity (psu) September 3 rd - September 16 th 2009 Week cbnc3

28 28 4-Week Objective: To view seasonal patterns for 2009 during periods of high and low discharge Graphs of Sandi, am169 –Smoothed to 1 week, original data Graphs of tides and discharge –Smoothed to 1 week

29 29 4-Week page sandi discharge Tide am169

30 30 4-Week Conclusions: Low River Discharge Difference between simulated and observed values is close to 0 psu during spring to neap transitions At am169, difference between simulated and observed values are up to 12 psu transitioning from neap to spring tides At Sandi, difference between simulated and observed values are up to 7 psu transitioning from neap to spring tides

31 31 4-Week Conclusions: High River Discharge During highest discharge: difference between simulated and observed values is consistent at ≈5 psu (Am169) or ≈2 psu (Sandi) Once discharge begins to drop transitional differences emerge Sandi Salinity (psu) Cubic meters/second May 21 st - June 17 th 2009

32 32 Annual Objective: To view long term trends between tides, discharge, wind direction and salinity values Graphs of Sandi and am169 –Smoothed to 28 days Graphs of tides, discharge and wind velocity –Smoothed to 28 days

33 33 Annual Page sandi discharge wind Tide

34 34 Annual Pages: Conclusions Sandi –Constant difference between simulated and observed values of 3-5 psu during the year –Salinity values for both simulated and observed drop when river discharge increase, and rise as river discharge drops am169 –Observed values show a monthly fluctuation that is not apparent in the simulated values –Salinity values for both simulated and observed drop when river discharge increase, and rise as discharge drops

35 35 Pros and Cons of Time Windows 2 weeks –Pros: see the small patterns and easy viewing –Cons: no long term trends 4 weeks –Pros: can see some long term trends, effects of discharge are more visible. –Cons: too crowded, only seasonal Annual –Pros: see long term trends –Cons: no short term trends, no slight fluctuations

36 36 Future Research Look at multiple years to find trends in simulated data versus observed data Repeat data analysis after changes to the model have been implemented El Niño and La Niña influence in past years Create plot configurations for temperature (2 weeks, 6 months, annual) Create plot configurations using 6 month segments instead of 4 week segments

37 37 Future Recommendations Create plot configurations for velocity data to see if similar trends to the salinity data exist

38 38 Future Recommendations Statistical analysis on simulated data and observed data Weeks Am169 index of agreement 0-1

39 39 Acknowledgments Pat Welle Dr.Antonio Baptista Dr. Grant Law Dr. Charles Seaton Karen Wegner Bonnie Gibbs Elizabeth Woody National Science Foundation Saturday Academy –Apprenticeship in Science and Engineering(ASE)

40 40 Thank you! Any Questions?


Download ppt "1 Understanding Salinity Variability in the Columbia River Estuary Sierra & Julia Observation ● Prediction ● Analysis ● Collaboration Frontline Mentor:"

Similar presentations


Ads by Google