As I've mentioned before, there was already a hydrodynamic with water temperature model written. Task 1 has been getting that model running for a twelve day simulation in August 2004. Initial results are so-so. The good news is the model runs, the bad news is it runs too hot! Here are some pictures to help you (and me) understand a little bit better. The first is a shot of the surface water temperature at the start of 12 day run. Ignore the date on the top of the figure, its wrong. It should be 8/7/04. I'll fix it later. Anyway, this figure shows a few things. First, you can see the domain I'm modeling! Isn't it pretty? Do you recognize this as the San Joaquin River and Stockton Deep Water Ship Channel? Maybe not. But, for those of you who may be familiar with Stockton (Katie, thats you!). All the way to the right of the picture, in the top corner, that's the Weber Point area. And on the bottom right, thats the San Joaquin River. My sister Katie used to live just to the right of this picture. The island area, between x = 300 and x = 500 is Rough and Ready Island. Anyway, in this snapshot, the surface temperature is the same everywhere. We did this on purpose. To start the model, we used an initial condition of 24.15 oC. This helps stabilize the model and keep the numbers out of control.
The next picture shows the surface water temperature 12 days (or 288 hours) later. You can see, things have changed, and the surface water temperature is no longer constant throughout the whole domain. Temperature can be added to the system two ways, the first is through the boundaries. We have four boundaries in our model. A boundary is basically a spot where we decide to end the model, but we still want stuff to happen at that spot, so we tell the model the information for that location. So, water from the bay (that boundary is the one to the left of the domain) and water from the San Joaquin River (bottom right of figure) bring into the domain heat. The second, and much more important source of heat is the exchange with the atmosphere. The sun heats up the water, wind can cause evaporation which cools the water, for example.
So, at least our model is doing something! It heats up and cools down each day. But, as the next plot shows, our model results don't quite resemble what we saw when we did our field work. Bummer. This plot is only for a 30 hour segment because that's the length of time we did observations for. But, you can clearly see that our model is too warm! There are a number of explanations for this. So, we're searching for the correct one. Right now we think our heat fluxes are too high. We estimate the effect the atmosphere and weather has on the temperature by using data collected on farms throughout California. Sometimes these are good estimates for lake and river areas, sometimes they are not. We have to try and explain why the data might not be appropriate for our system and what 'manipulations' we can make to the data to get the model to match our observations. Fun times.
So, at least our model is doing something! It heats up and cools down each day. But, as the next plot shows, our model results don't quite resemble what we saw when we did our field work. Bummer. This plot is only for a 30 hour segment because that's the length of time we did observations for. But, you can clearly see that our model is too warm! There are a number of explanations for this. So, we're searching for the correct one. Right now we think our heat fluxes are too high. We estimate the effect the atmosphere and weather has on the temperature by using data collected on farms throughout California. Sometimes these are good estimates for lake and river areas, sometimes they are not. We have to try and explain why the data might not be appropriate for our system and what 'manipulations' we can make to the data to get the model to match our observations. Fun times.
By the way, creating these plots is a 2-step process. The first is using Fortran to run my model, the 2nd is using a program called Matlab to create the plots. My model runs for 6 hours, then I can spend a few minutes creating these plots to see the results. Often, I leave two or three models running over night and come back and check the results.
While the hydrodynamic model runs, I have been working on getting MY water quality model running, and, today, I have been successful. I've been working towards this for three years, and I finally got all my equations together, compiled in Fortran and off it runs!!! Its a big step towards the end of my research. It feels great, but I'm, exhausted after a few days of chasing down bugs and fixes. . .
anyway, these next pics probably only excite me, but they are screenshots of my code and my model running. sweet!
And with that, I head to the bus. . . . 7 more days of work left here.
While the hydrodynamic model runs, I have been working on getting MY water quality model running, and, today, I have been successful. I've been working towards this for three years, and I finally got all my equations together, compiled in Fortran and off it runs!!! Its a big step towards the end of my research. It feels great, but I'm, exhausted after a few days of chasing down bugs and fixes. . .
anyway, these next pics probably only excite me, but they are screenshots of my code and my model running. sweet!
And with that, I head to the bus. . . . 7 more days of work left here.
2 comments:
Way to go Laura!!
Congratulations! You rock!!
Post a Comment