Short paper workplan

[cclatterbuck]

CCD = coastal cleanup day data OC = ocean conservancy data

Research questions & ideas:

1. What plastics are most commonly collected in California?

   • [x] Figure 1A: boxplot of the top 10-15 plastics categories based on mean annual counts, 1988-current
   • [x] Figure 1B: boxplot of the top 10-15 plastics categories based on normalized annual counts, 2016-current
     • See scripts/fig_coarse_boxplot.R and figures folder
   • [x] Table 1: Top 10-15 plastics categories which include: # of years collected, frequency on top x list, mean rank, summed counts, summed normalized counts
   • [ ] Appendix figures: for aggregate plastics categories that separate plastic vs. non-plastic items later in the collection period (e.g., bags --> plastic bags and paper bags), present figures demonstrating the percentage of plastic items vs. other items in that category. Preliminary figures already available in slides 7-11 of this slide deck.

2. Have the commonly collected plastic types changed over time?

   • [x] Figure 2A: heatmap of plastics categories ranked by annual counts over time (1988-current)
   • [x] Figure 2B: heatmap of plastics categories ranked by normalized annual counts over time (2016-current)
     • See scripts/fig_coarse_heatmap.R and figures folder
   • [ ] Test whether top x ranks have changed over time -- something like rank or mean order stability analysis? Ideally determine for both overall and normalized counts, if possible

3. Are there any hotspots of plastics collection? Are these hotspots racially equitable?

   • 2016-current data only
   • Helpful paper: Comparison of hotspot methodologies for survey data
   • Method selection: Getis-Ord-Gi, which requires amassing points into polygons. Need to think about what kinds of polygons makes the most sense here -- likely census districts? Also consider minimum # of sampling events to include.
   • Standardize data: See this paper section 2.3. I think the “summed effort-corrected counts” are equivalent to the total #s of plastics collected over number of people surveying; “mean effort-corrected counts” are equivalent to taking the summed effort-corrected counts over the number of trash events.
   • Will need to decide whether to limit to overall plastics or look at some of the highest plastics counts – e.g., cigarettes, bottles, etc. See section 2.4.4 in the same paper for hotspots conditional on plastic type presence (species presence).

4. Do commonly collected plastic types vary geographically:

   • Coastal vs inland counties?
   • GIS to determine other coastal vs. inland attributes (e.g., mileage from coast, etc.)
     • [x] Calculate distance from coast for 2016-current data in the .ipynb jupyter notebook
     • [ ] Regression

[cclatterbuck]

Additional analyses and visualizations:

• [ ] Timeseries of OC data, look for any increases or decreases in collected items over time (normalized data)
• [ ] Compare plastic type hotspots to CES 4.0 data -- Scott noted it may take time to access this data
• [ ] R Shiny bubble plot of total items & breakdown by major plastic types (normalized data)