Class on December 4, 2018
Faculty and students spent the class period working on capstone project planning and working on initial steps in the plan. Sub-activities included white board brainstorming and online document research.
Online documents are rich sources of finding useful values for parameters and coefficients required of food web models and NPZ models.
Lucie helped students perform online document research for the parameters and coefficients they needed (e.g. food web efficiencies). Bruce worked with the Ankobra NPZ modeling team to plan for their modeling effort. Their whiteboard session produced:
which suggested important modeling considerations to build upon are seasonal (dry v. wet season) differences in precipitation, turbidity, residence time, wind, and mixing/flushing. Methodologies to incorporate a time buffer for effects on phytoplankton behavior were discussed. Daylight hours per day and temperature are uniform year-round so would likely not be relevant to consider outside of setting informing fixed parameter values.
Some of the dry v. wet season differences are dramatic:
1. Secchi disk visibility is 1.4 of the average 8m depth in dry season compared to only 0.07m of the average 8m depth in wet season.
2. Average flow of the Ankobra river is effectively 0 in dry season compared to 0.3 meters-cubed per second in wet season.
3. Residence time is dramatically different based on the difference in flow.
4. Precipitation and Wind values are different as well.
The Ankobra team found the Elkhorn Slough in California to have conditions similar to the Ankobra, with the benefit of having 18 years worth of relevant turbidity and other information. In the absence of Ankobra historical data, the team decided to create a model for the Elkhorn estuary before transferring lessons learned to an Ankobra model.
As a result, the team spent the remainder of class getting downloaded data sets into a Python notebook analysis, looking at preliminary data patterns and trends:
Lucie helped students perform online document research for the parameters and coefficients they needed (e.g. food web efficiencies). Bruce worked with the Ankobra NPZ modeling team to plan for their modeling effort. Their whiteboard session produced:
which suggested important modeling considerations to build upon are seasonal (dry v. wet season) differences in precipitation, turbidity, residence time, wind, and mixing/flushing. Methodologies to incorporate a time buffer for effects on phytoplankton behavior were discussed. Daylight hours per day and temperature are uniform year-round so would likely not be relevant to consider outside of setting informing fixed parameter values.
Some of the dry v. wet season differences are dramatic:
1. Secchi disk visibility is 1.4 of the average 8m depth in dry season compared to only 0.07m of the average 8m depth in wet season.
2. Average flow of the Ankobra river is effectively 0 in dry season compared to 0.3 meters-cubed per second in wet season.
3. Residence time is dramatically different based on the difference in flow.
4. Precipitation and Wind values are different as well.
The Ankobra team found the Elkhorn Slough in California to have conditions similar to the Ankobra, with the benefit of having 18 years worth of relevant turbidity and other information. In the absence of Ankobra historical data, the team decided to create a model for the Elkhorn estuary before transferring lessons learned to an Ankobra model.
As a result, the team spent the remainder of class getting downloaded data sets into a Python notebook analysis, looking at preliminary data patterns and trends:
