You are here

Week 5: Astoria & Analyzing River/Ocean Influences On Data Differences

Monday: Full day of script changes and producing plots for other sensors. I had my hands full with plotting model and observational data for sensor locations SATURN-02 and Jetty A. The underlying process was essentially the same as I had been doing, but I had to alter parts of it where the model extraction was included. There was some redefining of variables and other Matlab script errors that I had to correct, but I was able to successfully conclude with temperature and salinity model-observation data comparison plots for both locations. Grant is also working on a short movie clip file for the Database 22 cut area (from mouth of Columbia River to Beaver Army) during the 10-day timeseries in May that I’ve been using. 
Tuesday: Our trip to Astoria was a good learning experience. Because Logan and I download observational data from the CMOP website, I found it quite useful to learn a little more about the sensors themselves and how data is transmitted to the stations. It was interesting to see the gliders used for data collection and then hear more about the technical maintenance of sensors in the network. Afterwards, we ate a picnic lunch at the Astoria Column then climbed our way up to the top of the column (easier said than done). Our next stop was the Maritime Museum. Here, we explored the different historical ships that set sail in the area, as well as the trading industry, and how Native Americans were involved. Overall, it was a nice drive and trip up north.
Wednesday, Thursday, Friday: After plotting the model-data comparisons, some time was spent on trying to figure out what was contributing to the error of the initial sensor SATURN-07. I found that the difference between model and observational data for SATURN-02 was very small compared to the data comparison of SATURN-07. The model was slightly lower than the observational data for both temperature and salinity, and the fluctuation was mirrored fairly well in the model as compared to the actual readings. I was very pleased with Jetty A’s model-data comparisons; both temperature and salinity matched up almost exactly with the actual data. The amount of fluctuation in Jetty A’s location was pretty high compared to other sites. After viewing the completed video for the salinity intrusion of the estuary, it became apparent that most of the ocean’s input entered into Baker Bay through a main channel upstream of Jetty A’s channel. It looked like the saltwater propogated through there from a high tide, spread through the bay then most of it exited through this same channel and back into the river when the tide went out. Because the SATURN-07 comparison illustrates an overshoot of salinity and cooler temperatures than the actual data, I figured that it could be too much of an incorporation of ocean input or too little of river discharge. After seeing that there is efficient modeling of SATURN-02, I figured that the saltwater values were pretty accurate, so it could be an inaccurate inclusion of river discharge. Joseph Zhang also gave an interesting suggestion after Grant and I informed him of the dilemma- he suggested that there may be a freshwater input that is not being accounted for in the grid, which would cause a decrease in salinity and warmer temperatures in the bay. He guessed that the Chinook River connecting to the northern boundary of Baker Bay could be this missing piece. And sure enough, after looking at a map, there is a river that dumps into the bay that is not a part of the model grid. We aren’t sure if this river plays an integral part of bay parameter levels and if it is significant enough to have an effect, however, so that is to be determined. The model’s lack of accuracy when it comes to including mixing processes could also very well play a role. This physical process is strengthened or weakened during spring and neap tides so my next step is to observe the spring and neap tide conditions. In order to do this, I need to plot data during the whole month of May, as it is on a lunar cycle. In this manner, I can identify the time intervals of spring and neap tides, and see if there is some correlation between modeling accuracy and precision and these tidal cycles. I can also ensure I have correct tidal cycle start and end dates by looking at the elevation plot for SATURN-07 (only have model data) and see where the fluctuation range is the largest (higher highs and lower lows) to find the spring tide and then smaller fluctuation (neap tide). Moji put the model data into my directory on Friday, so I will be working on the script for getting this data comparison plotted for SATURN-02 and Jetty A for the whole month of May, and then the rest of the model data for the month of May for SATURN-07 (only have observational data for 13 days).