Cleanup 2020 data

[cclatterbuck]

Currently using cleanups from the entire month of September 2020 as the month was designated "Coastal Cleanup Month" due to COVID-19. Two ways to reduce the additional cleanup events from this month:

• Look at Win Cowger's work re: cleanup efforts over the same area over a year and how collection data relates to land use, rain, etc.
• Define a buffer around each cleanup point for the month; for any overlapping buffers, retain the cleanup that occurred earliest.

[sydneyrilum]

KEY TAKEAWAYS:

• For Sept. 2020 cleanups, only use the first cleanup at each "site"
• Other studies have defined "unique" cleanup areas as sites located 100 m to 1000 m apart
• Outliers - only remove cleanups when necessary and make sure to document reasons for why certain ones were excluded

Mini literature review findings:

Cowger, W. et al. (2019) Anthropogenic litter cleanups in Iowa riparian areas reveal the importance of near-stream and watershed scale land use

• “To avoid complicating the analysis, we only used data from river reaches with no known previous cleanup and normalized the anthropogenic litter data by stream length. Three data points from the records between 2006 and 2016 were removed from the dataset because they overlapped with a previous cleanup or included trash mass from cleanups not conducted in the stream.”
• They determined EFFORT by multiplying the duration of each cleanup by the average number of daily volunteers to create a “Person Days” variable (something we can’t do unfortunately)
• Used NLCD (National Land Cover Dataset) from the USGS, contains raster image that identifies 16 land use types that were generalized into 3 land use categories -> Used Spearman correlation matrices to compare interactions between land use types and anthropogenic litter and found that land use features associated with higher human exposure (roads and developed areas) resulted in higher riparian litter concentrations
• NOTE: this study used trash mass instead of trash count

Cowger, W. et al. (2022) Litter origins, accumulation rates, and hierarchical composition on urban roadsides of the Inland Empire, California

• In Methods -> “surveyed a unique ~100-1000 m length of roadside for litter 1-3 times per week for 2-4 weeks”

Nelms, S.E. et al. (2017) Marine anthropogenic litter on British beaches: A 10-year nationwide assessment using citizen science data

• *a good paper to reference for publishing purposes…
• Used citizen science data from 2005-2014, beach litter surveys that gathered info on # of volunteers, duration of survey, distance covered; Allowed for easier data standardization; Typically, 100m+ long surveys
• Outliers: “Various methods of outlier removal were investigated but it was preferred that all data collected were utilised.”

Note: a lot of Win Cowger’s papers involve more standardized and experimental trash collection methods for the purpose of the study, not sourced from citizen science historical data collection

• used cleanup site length/area to be able to create a trash concentration value… (something we can’t do with only 1 lat/long per cleanup site record)
• In this Community Science-Informed Local Policy paper, they used the “Trash Monitoring Playbook” when surveying for trash, aka a robust standardized methodology, easy for comparison to other studies in CA
• Outliers: “Sites with fewer than 10 pieces of trash (n = 5) were not used for this part of the analysis as they may have added to the variability of the analysis.“

[cclatterbuck]

KEY TAKEAWAYS:

• For Sept. 2020 cleanups, only use the first cleanup at each "site"
• Other studies have defined "unique" cleanup areas as sites located 100 m to 1000 m apart
• Outliers - only remove cleanups when necessary and make sure to document reasons for why certain ones were excluded
  ...
• Outliers: “Various methods of outlier removal were investigated but it was preferred that all data collected were utilised.”
  ...
• Outliers: “Sites with fewer than 10 pieces of trash (n = 5) were not used for this part of the analysis as they may have added to the variability of the analysis.“

Thanks for the summary @sydneyrilum ! Just highlighting the findings above about outliers for issue #6