PRISM

[wbinzhe]

Dropbox of prism/annual

• data by var
• code: prism_annual.R
• output: zcta3_annual.csv https://www.dropbox.com/sh/mzc06a3x7psf4ch/AAAZ662Fibf5YtnfHIzm7Wzra?dl=0

[shoonlee]

@anishan

• Process county-level climate data. The key here is to spatially merge and assign a 4km by 4km grid to a county. Binzhe has already written a code to produce a similar product in a 3-digits zip code level and also created separate documentation for this procedure.
• Create a scatter plot using the county level temperature data comparing the average daily mean temperature between (1) 2006 and 2019 and (2) 2006-2008 and 2017-2019. Repeat the same exercise with the average daily maximum temperature. Here we want to see if there's a meaningful temperature change in the past 15 years (ideally for both hotter and colder days). We could also consider creating a map indicating counties that have become hotter vs colder. Anyhow, we want to identify if counties experience more extreme temperatures over the sample period.
• Create a histogram informing how many unique temperature bins a county has for a given year (try for 2006, 2012, and 2018). For instance, if a Cook county has a positive number of days in temperature bin 30-40, 40-50, 50-60, and 90+, it's going to be 4. Do this for every county and create a histogram. This is to validate the econometric model we're using.

[shoonlee]

@anishan

• Currently, the unit of the temperature is Celsius. Please convert it to the Fahrenheit and repeat the bucket process.
• The bucket data should be summarized at the year level, where each bucket has the number of days within a year that belongs to each bucket. So please create three different versions. The first two are what you've already created (although they have to be expressed in Fahrenheit). The third one is the bucket version aggregated at the annual level.
• Please start working on the descriptive analysis (creating figures, histograms that I described earlier).