HI_Climate_Data

Scripts to access and create climate futures from HI climate data sets

Climate Data sets

Climate data were provided by Dr. Abby Frazier in 2022. Descriptions of the data sets are provided below.

Zhang (dynamical)

• ncdf difficult to work with and don't behave like normal ncdf - can send monthly rasters
• name of years is wrong in historical (should be 1990-2009 but left as 2080-2099)
• raw nc
• processed
• For each island present (1990-2009), rcp45 and 85 (2080-2099)

Data prep scripts

Zhang extraction

Zhang_extraction.R

Extracts Zhang data and puts into RDS files. Requires external HD with Zhang files to run. Output goes to local park file.

Summarize rainfall observations

Mean_Rainfall_Obs.R -- Don't have to run again --

Takes data from Rainfall.Atlas and summarizes by year and month for bias correction process. Output is geoTIFF files to be read in to BC script.

Output are 1990-2009 tifts in RFMonthYr_Rasters_State_mm_1920_2012 folder

Bias correction

Bias_correction.R

Reads in OBS data and converts to RDS files. Obs data stored on external HD and output goes to local park file.

• Monthly temperature: Observation-based present day means use the period 1957-1981. For more information, please see the Climate of Hawaii website; Giambelluca et al. 2014; http://climate.geography.hawaii.edu/. Data are in degC

• Monthly precipitation: The observation-based present day mean maps use the period 1990-2009. These are calculated from the month-year rainfall maps, Frazier et al. 2016; available at http://rainfall.geography.hawaii.edu. Present-day means are calculated for each season: Annual, Wet Season (Nov-Apr), and Dry Season (May-Oct). Means are converted to different units: Rainfall is converted from mm to inches (divide mm by 25.4).

Variables created

• ANNPrecip_Obs (output are in mm)
• ANNPrecipDelta_rcp45/85
• ANNPrecipBC_rcp45/85...
  • JanPrecip
  • FebPrecip
  • MarPrecip...
  • WetPrecip
  • DryPrecip
• ANNTmean... (Output are in Fahrenheit)

Bias correction is Future delta (from Zhang extraction; downscaled future - downscaled obs) + Obs data

Wet/Dry season BC for plots

• Wet is Nov-Apr, Dry is May-Oct per Zhang README from Frazier
  • Precip is mean delta of each month

Variable creation

Variable_creation.R

Reads in bias-corrected data, calculates ts files for variables to be read into plotting scripts

• Output: Zone-Monthly-data.csv (Output are in Inches and Fahreneit)

SPI

SPI_analysis.R

Only SPI can be calculated b/c evapotranspiration calculations require more sophisticated methods than what can be done with T and P (Thornthwaite), thus ET cannot be calculated from available projection data.

Uses raw Zhang and Obs data, from Z drive because easiest to calculate on raster stack, rather than stars objects. SPI calculated on recent with whole period as reference. Future calculations use recent as reference period.

Plots average SPI for whole time period. Expect averages to be near 0.

Timeseries plots for each CF,

• Calculates SPI on stars objects for Obs and each CF.
• Save mean() stars objects for SPI
• Reduce ts for each climate zone and whole park
• Calculate characteristics for each climate zone
• Plot Maps w/ SPI ts for Obs and each CF
• Standard panel plot for each zone

Plotting scripts

Delta maps for whole park with ts avg of park

Maps_ts_plots.R

Plots map of whole park for each CF with ts plot that is average of the park. Input data are Mean RDS for whole park and annual timeseries

• Annual Tmean (ANNTmean)
• Annual Precip (ANNPrecip)

Topo clipped to HALE boundary. 100m Natural Earth from https://earthworks.stanford.edu/catalog/stanford-qd855xh2430

Wet/Dry seaonal maps and dotplots by zone

Maps_seasonal_dotplots.R

• Delta_monthly_plots to get seasonal delta csv
• Wet/Dry Tmean/Precip (WetTmean) (DryPrecip)