Data Sourcing and Processing

For monthly temperature data, this file shows:

data sourcing and processing
exploration of summary statistics at:
- full history
- monthly level
- seasonal level
- yearly level
outputting of raw data for sonification

Code

# Set up packages and functions
library(tidyverse)
library(kableExtra)
library(htmlTable)
library(gt)
source("code/cummax_ignore_na.R")
source("code/cummin_ignore_na.R")

Data Source

Three data sets were downloaded from the NOAA National Weather service NOWData tool.

All data sets are for Paso Robles, CA, from the start of data collection (“por”) to present (2025).

NOAA NOW Data map tool visualizing location of data source in Paso Robles, CA

We requested the monthly maximums, monthly minimums, and monthly averages.

Screenshot of inputs for the monthly maximums data set After clicking “Go”, a data table pops up. We copied it and pasted into Microsoft Excel, removed the summary rows included at the bottom, then saved as a csv.

This process yielded 3 csv files: paso_max.csv, paso_min.csv, and paso_avg.csv.

Process Data

Read in, combine, and process data from the csv files obtained above.

Code

## Read in Csv's for min, mean, max
min_data <- read.csv("data/paso_min.csv", na.strings = "M")
min_data <- min_data %>% 
  select(-Annual) %>% 
  pivot_longer( ## one row = one month
    cols = Jan:Dec, 
    names_to ="month", 
    values_to = "min"
    )



mean_data <- read.csv("data/paso_avg.csv", na.strings = "M")

mean_data <- mean_data %>% 
  select(-Annual) %>% 
  pivot_longer( ## one row = one month
    cols = Jan:Dec, 
    names_to ="month", 
    values_to = "mean"
    )


max_data <- read.csv("data/paso_max.csv", na.strings = "M")

max_data <- max_data %>% 
  select(-Annual) %>% 
  pivot_longer( ## one row = one month
    cols = Jan:Dec,
    names_to ="month", 
    values_to = "max"
    )

## combine data sets and add date formatting
monthly_weather <- bind_cols(min_data, 
                             mean_data[,3], 
                             max_data[,3]) %>%
  mutate(date = ym(paste0(year, "-", month))) 

## add seasonal variables 

monthly_weather <- monthly_weather %>% 
  mutate(
    season = case_when(
      month %in% c("Dec", "Jan", "Feb") ~ "Winter",
      month %in% c("Mar", "Apr", "May")  ~ "Spring",
      month %in% c("June", "July", "Aug")  ~ "Summer",
      month %in% c("Sept", "Oct", "Nov") ~ "Fall"
  ),
  season_color = case_when(
      month %in%  c("Dec", "Jan", "Feb")  ~ "#2f77c3",
      month %in% c("Mar", "Apr", "May")  ~ "#61bf9a",
      month %in% c("June", "July", "Aug")  ~ "#f94994",
      month %in% c("Sept", "Oct", "Nov") ~ "#eb9911"
  ), 
  season_year = case_when(
      month == "Dec" ~ year + 1,  # December belongs to *next* Jan/Feb
      .default = year
  ),
    season_label = paste(season, season_year)
  ) %>%   
  group_by(season, year) %>%
  ungroup()  %>%
  mutate(
         month = factor(month, levels = c("Dec", "Jan", "Feb", "Mar", "Apr", "May" , "June", "July", "Aug",  "Sept", "Oct",  "Nov") ))



monthly_weather <- monthly_weather %>%
  group_by(season_label, season_year) %>%
  mutate(
    xmin = min(date),
    xmax = max(date),
    seas_avg = mean(mean),
    seas_max = max(max),
    seas_min = min(min),
    season_color = unique(season_color),
    season = unique(season)
  ) %>%
  ungroup() %>%
  mutate(season = factor(season, levels = c("Winter", "Spring", "Summer", "Fall")))

This gets us three monthly time series, which we can visualize.

Visualize Raw Data

Code

## color palatte
plot_cols <- c("max" = "red", "min" = "blue", "mean"="grey", "Winter" = "#2f77c3", "Spring" = "#61bf9a", "Summer" = "#f94994", "Fall" =  "#eb9911")

monthly_weather %>% 
  ggplot(aes(x = date)) +
  geom_line(aes(y = min, col = "min")) + 
  geom_line(aes(y = max, col = "max")) + 
  geom_line(aes(y = mean, col = "mean")) + 
  scale_color_manual(values = plot_cols)

We can clearly see that the maximum series is consistently above the average, which is consistently above the minimum, as we expect.

It’s hard to see the monthly variation on this long time scale, so we will zoom in to just 10 years of data starting after the missing data period at the beginning (after 1909):.

Zoom in

Code

monthly_weather %>% 
  filter(between(year, 1910, 1919)) %>%
  ggplot(aes(x = date)) +
  geom_line(aes(y = min, col = "min")) + 
  geom_point(aes(y = min, col = "min"))+
  geom_line(aes(y = max, col = "max")) + 
  geom_point(aes(y = max, col = "max"))+ 
  geom_line(aes(y = mean, col = "mean")) +  
  geom_point(aes(y = mean, col = "mean"))+
  scale_color_manual(values = plot_cols)

Each dot on the plot corresponds to one month in a given year. The annual seasonal patterns are defined by the apparent winter valleys and summer peaks.

Seasonal Info

We can add the information about each season by coloring the points in the above plot:

Code

monthly_weather %>% 
  filter(between(year, 1910, 1919)) %>%
  ggplot(aes(x = date)) +
  geom_line(aes(y = min, col = "min")) + 
  geom_point(aes(y = min, col = season))+
  geom_line(aes(y = max, col = "max")) + 
  geom_point(aes(y = max, col = season))+ 
  geom_line(aes(y = mean, col = "mean")) +  
  geom_point(aes(y = mean, col = season))+
  scale_color_manual(values = plot_cols)

Summer and winter are confirmed as the high and low points, and we see spring and fall on the appropriate “sides” of the distribution. We see these relative seasonal patterns in all three summary statistics, which makes sense: we expect the lowest high to be in winter, and the highest low to be in summer, and so on.

Since months and seasons both capture the structure of the annual variation, we can aggregate to the monthly or seasonal level to gain additional insight to the series as a whole (without having to zoom in).

Summary Statistics

We will visually (plots) and numerically (tables) examine the following statistics:

minimum of monthly mins
mean of monthly means
max of monthly maxxes

All Time Summary Statistics

For the full history January 1900- May 2025.

Code

summary_table_all <- monthly_weather %>% 
  summarise(
    ## mean 
    mean_Min = mean(min, na.rm = T), 
    mean_Mean = mean(mean, na.rm = T), 
    mean_Max = mean(max, na.rm = T), 
    ## sd
    sd_Min= sd(min, na.rm = T),
    sd_Mean = sd(mean, na.rm = T),
    sd_Max = sd(max, na.rm = T),
    ## max 
    max_Min = max(min, na.rm = T),
    max_Mean = max(mean, na.rm = T),
    max_Max = max(max, na.rm = T),
    ## min
    min_Min = min(min, na.rm = T),
    min_Mean = min(mean, na.rm = T),
    min_Max = min(max, na.rm = T),
  )

## chat help

facet_vars <- c("min", "mean", "max")

# Step 1: Pivot your summary table
vline_data <- summary_table_all %>%
  pivot_longer(cols = contains(c("Min_min", "Mean_mean", "Max_max")),
               names_to = "name", values_to = "xintercept") 
vline_data$name = facet_vars


monthly_weather %>%
  bind_cols(summary_table_all[rep(1, nrow(monthly_weather)), ]) %>%
  select(-c(xmin, xmax)) %>%
  pivot_longer(cols = contains(c("min", "mean", "max"))) %>%
  filter(name %in% facet_vars) %>%
  select(year, month, date, season, name, value) %>%
  ggplot(aes(x = value, fill = name)) +
  geom_histogram(bins = 30) +
  geom_vline(
    data = vline_data,
    aes(xintercept = xintercept),
    color = "black", linewidth = 2, linetype = "solid"
  ) +
  geom_vline(
    data = vline_data,
    aes(xintercept = xintercept, color = name),
    linewidth = 1, linetype = "solid"
  ) +
  facet_grid(name~., scales = "fixed") +
  theme_minimal() +
  scale_fill_manual(values = plot_cols)+
  scale_color_manual(values = plot_cols)

Code

## chat help

facet_vars <- c("min", "mean", "max")

# Step 1: Pivot your summary table
vline_data <- summary_table_all %>%
  pivot_longer(cols = contains(c("min_Min", "Mean_mean", "Max_max")),
               names_to = "name", values_to = "xintercept") 
vline_data$name = facet_vars


monthly_weather %>%
  bind_cols(summary_table_all[rep(1, nrow(monthly_weather)), ]) %>%
  select(-c(xmin, xmax)) %>%
  pivot_longer(cols = contains(c("min", "mean", "max"))) %>%
  filter(name %in% facet_vars) %>%
  select(year, month, date, season, name, value) %>%
  ggplot(aes(x = value, fill = name)) +
  geom_boxplot() +
    geom_vline(
    data = vline_data,
    aes(xintercept = xintercept),
    color = "black", linewidth = 2, linetype = "solid"
  ) +
  geom_vline(
    data = vline_data,
    aes(xintercept = xintercept, color = name),
    linewidth = 1, linetype = "solid"
  ) +
  facet_grid(name~., scales = "fixed") +
  theme_minimal() +
  scale_fill_manual(values = plot_cols)+
  scale_color_manual(values = plot_cols)

Code

summary_table_all %>%
select(-c(max_Min, min_Max, min_Mean, max_Mean))%>%
gt() %>%
  tab_header(
    title = "All-time Summaries"
  ) %>%
  fmt_number(
    columns = where(is.numeric), 
    decimals = 2
  )%>%
  cols_label(
    min_Min = "Min", 
    mean_Min = "Mean",
    sd_Min = "SD", 
    mean_Mean = "Mean", 
    sd_Mean = "SD", 
    max_Max = "Max",
    mean_Max = "Mean",
    sd_Max= "SD"
  ) %>%
  tab_spanner(
    label = "Monthly Minimums",
    columns = c(min_Min, mean_Min, sd_Min)
  ) %>%
  tab_spanner(
    label = "Monthly Means",
    columns = c(mean_Mean, sd_Mean)
  ) %>%
  tab_spanner(
    label = "Monthly Maximums",
    columns = c(mean_Max, sd_Max, max_Max)
)

All-time Summaries
Monthly Means		Monthly Maximums			Monthly Minimums
Mean	SD	Mean	SD	Max	Min	Mean	SD
59.02	9.22	90.69	13.27	117.00	0.00	32.13	8.52

Monthly Level Summaries

Code

month.abb = c("Dec", "Jan", "Feb", "Mar", "Apr", "May" , "June", "July", "Aug",  "Sept", "Oct",  "Nov") 
summary_table <- monthly_weather %>% 
  mutate(month = factor(month, levels = month.abb)) %>%
  group_by(month) %>%
   summarise(
    ## mean 
    mean_Min = mean(min, na.rm = T), 
    mean_Mean = mean(mean, na.rm = T), 
    mean_Max = mean(max, na.rm = T), 
    ## sd
    sd_Min= sd(min, na.rm = T),
    sd_Mean = sd(mean, na.rm = T),
    sd_Max = sd(max, na.rm = T),
    ## max 
    max_Min = max(min, na.rm = T),
    max_Mean = max(mean, na.rm = T),
    max_Max = max(max, na.rm = T),
    ## min
    min_Min = min(min, na.rm = T),
    min_Mean = min(mean, na.rm = T),
    min_Max = min(max, na.rm = T),
  )%>%
  arrange(month)

plot_cols <- c(plot_cols, "min_Min" = "blue", "mean_Mean" = "grey", "max_Max" = "red")

agg_plot<- summary_table %>% 
  ggplot(aes(x = month)) + 
  geom_line(aes(y = min_Min, col = "min_Min", group = 1), linewidth = .75) + 
  geom_line(aes(y = mean_Mean, col = "mean_Mean", group = 1), linewidth = .75) + 
  geom_line(aes(y = max_Max, col = "max_Max", group = 1), linewidth = .75) + 
  geom_point(aes(y = min_Min, col = "min_Min", group = 1), size = 3) + 
  geom_point(aes(y = mean_Mean, col = "mean_Mean", group = 1), size = 3) + 
  geom_point(aes(y = max_Max, col = "max_Max", group = 1), size = 3) + 
  scale_color_manual(values = plot_cols, name = "Legend") + 
  ylab("Temperature (F)")+ xlab("Month") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5)) + ylim(c(0,120))

agg_plot

Code

agg_plot + 
  geom_point(data = monthly_weather, aes(x = month, y = min, col = season), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = month, y = mean, col = season), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = month, y = max, col = season), alpha = 0.1, position = position_jitter())  +
  ylab("Temperature (F)")+ xlab("Month") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5)) + 
  ylim(c(0,120))

Code

summary_table %>%
select(-c(max_Min, min_Max, min_Mean, max_Mean))%>%
gt() %>%
  tab_header(
    title = "All-time Summaries"
  ) %>%
  fmt_number(
    columns = where(is.numeric), 
    decimals = 2
  )%>%
  cols_label(
    min_Min = "Min", 
    mean_Min = "Mean",
    sd_Min = "SD", 
    mean_Mean = "Mean", 
    sd_Mean = "SD", 
    max_Max = "Max",
    mean_Max = "Mean",
    sd_Max= "SD"
  ) %>%
  tab_spanner(
    label = "Monthly Minimums",
    columns = c(min_Min, mean_Min, sd_Min)
  ) %>%
  tab_spanner(
    label = "Monthly Means",
    columns = c(mean_Mean, sd_Mean)
  ) %>%
  tab_spanner(
    label = "Monthly Maximums",
    columns = c(mean_Max, sd_Max, max_Max)
)

All-time Summaries
month	Monthly Means		Monthly Maximums			Monthly Minimums
month	Mean	SD	Mean	SD	Max	Min	Mean	SD
Dec	46.95	2.47	73.08	4.52	87.00	7.00	21.21	4.44
Jan	47.05	2.71	72.75	4.51	83.00	0.00	21.74	5.42
Feb	50.06	2.68	75.99	4.91	85.00	13.00	25.34	4.28
Mar	53.06	2.79	81.32	5.29	91.00	20.00	29.23	3.69
Apr	56.75	2.63	89.06	5.51	100.00	24.00	31.55	3.29
May	61.84	2.53	96.31	4.90	110.00	30.00	35.67	3.34
June	67.56	2.52	103.65	4.09	115.00	31.00	39.69	3.40
July	71.77	2.36	106.84	3.44	115.00	36.00	43.12	3.33
Aug	71.47	2.14	105.86	3.35	117.00	32.00	42.58	3.66
Sept	68.33	2.48	103.47	4.54	115.00	28.00	39.08	3.78
Oct	61.41	2.50	96.05	4.43	108.00	19.00	31.71	4.30
Nov	52.72	2.39	84.25	4.96	95.00	14.00	24.86	4.47

Seasonal Level Summaries

Code

summary_table_seasonal <- monthly_weather %>% 
  mutate(season = factor(season, levels = c("Winter", "Spring", "Summer", "Fall"))) %>%
  mutate(month = factor(month, levels = month.abb)) %>%
  group_by(season) %>%
   summarise(
    ## mean 
    mean_Min = mean(min, na.rm = T), 
    mean_Mean = mean(mean, na.rm = T), 
    mean_Max = mean(max, na.rm = T), 
    ## sd
    sd_Min= sd(min, na.rm = T),
    sd_Mean = sd(mean, na.rm = T),
    sd_Max = sd(max, na.rm = T),
    ## max 
    max_Min = max(min, na.rm = T),
    max_Mean = max(mean, na.rm = T),
    max_Max = max(max, na.rm = T),
    ## min
    min_Min = min(min, na.rm = T),
    min_Mean = min(mean, na.rm = T),
    min_Max = min(max, na.rm = T),
  )%>%
  arrange(season)

plot_cols <- c(plot_cols, "min_Min" = "blue", "mean_Mean" = "grey", "max_Max" = "red")



summary_table_seasonal %>% 
  ggplot(aes(x = season)) + 
  geom_line(aes(y = min_Min, col = "min_Min", group = 1), linewidth = .75) + 
  geom_line(aes(y = mean_Mean, col = "Mean_mean", group = 1), linewidth = .75) + 
  geom_line(aes(y = max_Max, col = "max_Max", group = 1), linewidth = .75) + 
  geom_point(aes(y = min_Min, col = season, group = 1), size = 3) + 
  geom_point(aes(y = mean_Mean, col = season, group = 1), size = 3) + 
  geom_point(aes(y = max_Max, col = season, group = 1), size = 3) + 
  scale_color_manual(values = plot_cols, name = "Legend") + 
  ylab("Temperature (F)")+ xlab("Season") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5)) + ylim(c(0,120))

Code

summary_table_seasonal %>% 
  ggplot(aes(x = season)) + 
  geom_line(aes(y = min_Min, col = "min_Min", group = 1), linewidth = .75) + 
  geom_line(aes(y = mean_Mean, col = "mean_Mean", group = 1), linewidth = .75) + 
  geom_line(aes(y = max_Max, col = "max_Max", group = 1), linewidth = .75) + 
  geom_point(aes(y = min_Min, col = season, group = 1), size = 3) + 
  geom_point(aes(y = mean_Mean, col = season, group = 1), size = 3) + 
  geom_point(aes(y = max_Max, col = season, group = 1), size = 3) + 
  scale_color_manual(values = plot_cols, name = "Legend") + 
  geom_point(data = monthly_weather, aes(x = season, y = min, col = season), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = season, y = mean, col = season), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = season, y = max, col = season), alpha = 0.1, position = position_jitter()) +
  ylab("Temperature (F)")+ xlab("Year") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5))+ ylim(c(0, 120))

Code

## chat help

facet_vars <- c("min", "mean", "max")

# Step 1: Pivot your summary table
vline_data <- summary_table_seasonal%>%
  group_by(season) %>%
  pivot_longer(cols = contains(c("min_Min", "mean_Mean", "max_Max")),
               names_to = "Statistic", values_to = "xintercept") 
vline_data$Statistic = rep(facet_vars, times = 4)


monthly_weather %>%
  bind_cols(summary_table_seasonal[rep(1, nrow(monthly_weather)), ]) %>%
  select(-c(xmin, xmax)) %>%
  pivot_longer(cols = contains(c("min", "mean", "max")), names_to = "Statistic") %>%
  filter(Statistic %in% facet_vars) %>%
  select(year, month, date, season...7, Statistic, value) %>%
  rename(season=season...7) %>%
  ggplot(aes(x = value, fill = season)) +
  geom_histogram(bins = 30) +
  geom_vline(
    data = vline_data,
    aes(xintercept = xintercept, col = Statistic),
    linewidth = 1, linetype = "dashed"
  ) +
  facet_grid(Statistic~season, scales = "fixed") +
  theme_minimal() +
  scale_fill_manual(values = plot_cols) + 
  scale_color_manual(values = plot_cols)

Code

summary_table_seasonal %>%
 select(-c(max_Min, min_Max, min_Mean, max_Mean))%>%
gt() %>%
  tab_header(
    title = "All-time Summaries"
  ) %>%
  fmt_number(
    columns = where(is.numeric), 
    decimals = 2
  )%>%
  cols_label(
    min_Min = "Min", 
    mean_Min = "Mean",
    sd_Min = "SD", 
    mean_Mean = "Mean", 
    sd_Mean = "SD", 
    max_Max = "Max",
    mean_Max = "Mean",
    sd_Max= "SD"
  ) %>%
  tab_spanner(
    label = "Monthly Minimums",
    columns = c(min_Min, mean_Min, sd_Min)
  ) %>%
  tab_spanner(
    label = "Monthly Means",
    columns = c(mean_Mean, sd_Mean)
  ) %>%
  tab_spanner(
    label = "Monthly Maximums",
    columns = c(mean_Max, sd_Max, max_Max)
)

All-time Summaries
season	Monthly Means		Monthly Maximums			Monthly Minimums
season	Mean	SD	Mean	SD	Max	Min	Mean	SD
Winter	48.03	2.99	73.94	4.86	87.00	0.00	22.77	5.07
Spring	57.19	4.48	88.89	8.05	110.00	20.00	32.15	4.35
Summer	70.27	3.02	105.45	3.87	117.00	31.00	41.79	3.77
Fall	60.82	6.87	94.62	9.17	115.00	14.00	31.90	7.16

Yearly

Code

summary_table_yearly<- monthly_weather %>% 
  group_by(year) %>%
   summarise(
    ## mean 
    mean_Min = mean(min, na.rm = T), 
    mean_Mean = mean(mean, na.rm = T), 
    mean_Max = mean(max, na.rm = T), 
    ## sd
    sd_Min= sd(min, na.rm = T),
    sd_Mean = sd(mean, na.rm = T),
    sd_Max = sd(max, na.rm = T),
    ## max 
    max_Min = max(min, na.rm = T),
    max_Mean = max(mean, na.rm = T),
    max_Max = max(max, na.rm = T),
    ## min
    min_Min = min(min, na.rm = T),
    min_Mean = min(mean, na.rm = T),
    min_Max = min(max, na.rm = T),
  )%>%
  arrange(year)

plot_cols <- c(plot_cols, "min_Min" = "blue", "mean_Mean" = "grey", "max_Max" = "red")



agg_plot <- summary_table_yearly %>% 
  filter(year >1909) %>%
  ggplot(aes(x = year)) + 
  geom_line(aes(y = min_Min, col = "min_Min", group = 1), linewidth = .75) + 
  geom_line(aes(y = mean_Mean, col = "Mean_mean", group = 1), linewidth = .75) + 
  geom_line(aes(y = max_Max, col = "max_Max", group = 1), linewidth = .75) + 
  geom_point(aes(y = min_Min, col = "min_Min", group = 1), size = 3) + 
  geom_point(aes(y = mean_Mean, col = "Mean_mean", group = 1), size = 3) + 
  geom_point(aes(y = max_Max, col = "max_Max", group = 1), size = 3) + 
  scale_color_manual(values = plot_cols, name = "Legend") + 
  ylab("Temperature (F)")+ xlab("Year") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5)) + ylim(c(0,120))

agg_plot

Code

agg_plot + 
  geom_point(data = monthly_weather, aes(x = year, y = min, col = "min"), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = year, y = mean, col = "mean"), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = year, y = max, col = "max"), alpha = 0.1, position = position_jitter()) +
  ylab("Temperature (F)")+ xlab("Year") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5))+ ylim(c(0, 120))

Code

agg_plot + 
  geom_point(data = monthly_weather, aes(x = year, y = min, col = season), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = year, y = mean, col =season), alpha = 0.1, position = position_jitter()) +
  geom_point(data = monthly_weather, aes(x = year, y = max, col = season), alpha = 0.1, position = position_jitter()) +
  ylab("Temperature (F)")+ xlab("Year") +
   theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5))+ ylim(c(0, 120))

Code

summary_table_yearly %>%
  mutate(year = ymd(paste0(year, "-01-01")))%>% select(-c(max_Min, min_Max, min_Mean, max_Mean))%>%
gt() %>%
  tab_header(
    title = "All-time Summaries"
  ) %>%
    fmt_number(
    columns = where(is.numeric), 
    decimals = 2
  )%>%
  fmt_date(
    columns = year,
    date_style = "year"
  )%>%
  cols_label(
    min_Min = "Min", 
    mean_Min = "Mean",
    sd_Min = "SD", 
    mean_Mean = "Mean", 
    sd_Mean = "SD", 
    max_Max = "Max",
    mean_Max = "Mean",
    sd_Max= "SD"
  ) %>%
  tab_spanner(
    label = "Monthly Minimums",
    columns = c(min_Min, mean_Min, sd_Min)
  ) %>%
  tab_spanner(
    label = "Monthly Means",
    columns = c(mean_Mean, sd_Mean)
  ) %>%
  tab_spanner(
    label = "Monthly Maximums",
    columns = c(mean_Max, sd_Max, max_Max)
)

All-time Summaries
year	Monthly Means		Monthly Maximums			Monthly Minimums
year	Mean	SD	Mean	SD	Max	Min	Mean	SD
1900	58.77	8.01	89.50	13.22	110.00	18.00	33.00	8.24
1901	58.44	9.56	89.00	15.30	110.00	16.00	31.17	8.71
1902	54.15	9.50	86.83	15.88	109.00	18.00	28.58	7.08
1903	54.67	10.53	85.43	17.12	106.00	15.00	27.14	8.28
1904	NaN	NA	NaN	NA	−Inf	Inf	NaN	NA
1905	NaN	NA	NaN	NA	−Inf	Inf	NaN	NA
1906	NaN	NA	NaN	NA	−Inf	Inf	NaN	NA
1907	NaN	NA	NaN	NA	−Inf	Inf	NaN	NA
1908	NaN	NA	NaN	NA	−Inf	Inf	NaN	NA
1909	62.28	9.91	95.40	16.35	106.00	15.00	29.00	8.37
1910	58.51	8.41	94.83	14.06	110.00	16.00	28.42	7.99
1911	56.71	8.74	90.83	13.49	110.00	13.00	28.58	8.77
1912	56.86	8.61	88.67	14.58	107.00	11.00	27.33	9.18
1913	58.06	9.85	93.08	14.08	112.00	0.00	27.17	10.62
1914	58.56	7.45	91.50	13.61	108.00	19.00	30.33	6.58
1915	58.77	8.86	91.58	14.76	109.00	17.00	29.92	8.07
1916	57.80	10.24	90.25	15.69	112.00	13.00	27.50	9.29
1917	58.87	11.25	94.00	15.83	115.00	18.00	28.58	8.13
1918	58.49	10.57	92.50	12.08	110.00	16.00	30.25	9.51
1919	57.98	9.92	90.33	15.26	112.00	16.00	29.42	8.13
1920	57.73	9.14	88.67	15.80	112.00	22.00	33.75	7.03
1921	58.90	9.06	91.75	12.64	111.00	19.00	32.58	7.76
1922	57.67	10.96	86.58	16.36	107.00	17.00	30.83	7.69
1923	57.87	9.34	90.25	14.04	109.00	17.00	32.50	8.47
1924	58.18	8.74	90.08	13.61	107.00	9.00	31.00	9.50
1925	58.41	8.00	90.83	14.17	115.00	24.00	33.00	6.58
1926	60.21	9.04	91.83	12.90	111.00	24.00	35.08	5.55
1927	58.90	8.68	89.67	13.83	106.00	24.00	33.75	6.08
1928	59.61	8.99	91.83	15.09	110.00	21.00	33.50	7.54
1929	59.01	10.11	91.83	15.07	111.00	21.00	32.25	8.01
1930	58.64	8.75	90.50	12.82	112.00	20.00	33.42	7.04
1931	60.32	10.43	92.50	14.62	113.00	20.00	34.58	7.51
1932	58.08	10.51	91.00	15.50	110.00	19.00	32.58	7.50
1933	57.62	10.34	91.42	17.55	117.00	21.00	31.25	7.84
1934	59.38	8.46	91.25	13.88	107.00	26.00	33.67	4.72
1935	57.05	10.23	88.92	13.81	109.00	22.00	31.33	7.64
1936	59.06	10.14	89.75	15.60	108.00	22.00	33.17	7.07
1937	58.11	11.21	88.67	16.07	108.00	14.00	32.50	8.82
1938	58.19	9.33	88.50	14.57	109.00	19.00	32.67	8.80
1939	59.02	9.81	90.75	14.08	107.00	23.00	31.50	6.60
1940	60.33	7.93	91.83	14.10	110.00	16.00	31.67	9.31
1941	60.15	7.14	89.08	12.69	108.00	18.00	32.17	7.93
1942	59.01	10.14	91.58	14.54	112.00	18.00	30.83	9.07
1943	59.49	7.81	92.33	11.97	108.00	15.00	31.25	8.02
1944	58.53	8.65	90.67	14.39	112.00	23.00	30.75	6.65
1945	58.66	9.54	91.67	14.36	109.00	16.00	29.33	8.26
1946	58.10	10.10	89.92	13.83	110.00	16.00	29.25	9.36
1947	59.31	10.12	93.67	13.58	110.00	15.00	29.75	9.24
1948	57.59	9.64	90.42	13.67	109.00	15.00	27.67	10.33
1949	57.54	10.69	89.50	15.71	109.00	10.00	29.25	10.22
1950	59.83	8.01	92.92	13.89	110.00	13.00	30.75	8.79
1951	58.34	9.29	90.67	14.69	106.00	21.00	30.42	8.07
1952	59.19	9.89	91.17	12.68	105.00	21.00	31.92	7.04
1953	58.98	8.78	91.50	10.47	106.00	16.00	29.67	9.27
1954	58.98	9.01	91.42	13.45	109.00	17.00	29.33	8.35
1955	57.89	9.51	90.17	13.32	111.00	19.00	29.42	7.14
1956	58.28	9.22	89.58	12.86	108.00	16.00	30.42	8.69
1957	59.27	9.89	89.67	14.94	111.00	17.00	32.00	10.54
1958	60.59	9.40	90.67	13.74	109.00	14.00	32.42	10.63
1959	60.41	9.41	92.58	11.97	112.00	17.00	30.58	9.08
1960	59.37	9.75	91.08	14.61	113.00	11.00	31.17	9.29
1961	59.73	10.54	91.08	14.44	112.00	17.00	30.00	9.53
1962	58.95	8.86	90.50	10.38	104.00	17.00	30.67	9.38
1963	58.55	9.30	88.25	11.99	104.00	12.00	31.50	9.45
1964	58.03	9.63	90.50	12.63	107.00	20.00	29.67	8.77
1965	57.69	9.08	88.42	9.97	101.00	19.00	30.67	7.49
1966	58.91	9.52	91.00	13.45	105.00	20.00	29.67	7.33
1967	59.20	10.94	88.67	14.37	106.00	15.00	32.50	10.02
1968	58.88	9.08	90.17	11.75	104.00	16.00	29.75	8.14
1969	59.07	9.66	89.50	13.55	108.00	19.00	32.58	8.44
1970	59.28	8.86	91.58	14.50	108.00	14.00	30.25	8.28
1971	57.92	10.25	90.58	13.43	110.00	19.00	28.17	8.90
1972	58.56	9.71	89.92	13.33	110.00	16.00	30.00	8.55
1973	57.20	9.13	88.33	15.29	106.00	15.00	29.55	7.06
1974	59.05	9.67	89.92	13.62	105.00	15.00	31.50	8.96
1975	57.81	9.24	90.50	12.00	109.00	14.00	28.83	10.25
1976	59.21	9.47	94.08	12.00	109.00	11.00	29.50	10.66
1977	59.58	9.23	91.08	11.83	105.00	19.00	30.25	8.16
1978	59.79	9.68	90.00	13.90	105.00	15.00	32.58	7.55
1979	60.17	10.13	89.50	15.01	109.00	19.00	31.33	9.24
1980	60.30	8.05	93.25	12.61	109.00	20.00	31.83	7.49
1981	61.16	8.77	90.42	12.16	106.00	25.00	34.17	6.99
1982	58.64	9.48	87.67	14.25	107.00	23.00	32.92	6.50
1983	60.61	8.95	88.25	14.55	109.00	24.00	36.17	7.49
1984	60.81	10.51	91.33	14.08	110.00	23.00	32.33	6.65
1985	56.78	9.66	90.75	12.84	107.00	20.00	31.42	8.58
1986	58.83	9.25	90.67	10.05	103.00	22.00	32.75	6.89
1987	61.10	7.77	89.92	14.61	105.00	14.00	31.75	10.96
1988	49.23	5.70	91.83	11.88	108.00	23.00	32.67	7.94
1989	60.33	9.45	92.00	11.62	110.00	18.00	32.50	9.43
1990	59.42	11.32	93.58	12.35	109.00	7.00	31.50	12.67
1991	60.83	8.79	90.92	13.06	108.00	17.00	32.58	9.90
1992	60.51	9.95	88.50	12.69	108.00	22.00	36.58	8.99
1993	59.22	8.77	89.83	13.08	107.00	20.00	33.58	8.38
1994	58.43	9.93	89.42	12.09	108.00	21.00	33.42	8.87
1995	60.17	7.98	88.50	12.31	105.00	30.00	35.67	5.55
1996	60.13	8.56	92.00	13.81	109.00	24.00	33.42	6.72
1997	59.90	9.01	89.42	12.71	108.00	21.00	34.42	8.30
1998	57.75	9.82	86.17	14.97	111.00	16.00	35.67	9.39
1999	57.55	8.89	88.67	12.12	104.00	21.00	33.42	8.34
2000	58.88	9.14	89.33	15.06	110.00	21.00	34.67	8.72
2001	59.16	10.25	90.83	14.35	110.00	22.00	35.08	8.75
2002	58.24	9.71	91.00	14.59	108.00	22.00	33.67	8.24
2003	58.31	9.60	87.09	13.90	105.00	21.00	33.55	8.77
2004	57.79	9.23	87.36	13.89	103.00	23.00	33.73	8.30
2005	59.71	9.11	87.58	12.40	105.00	26.00	35.92	6.75
2006	59.05	11.02	90.08	13.71	114.00	22.00	35.25	9.67
2007	59.59	10.53	93.83	11.87	108.00	13.00	33.58	10.59
2008	60.04	10.55	93.58	14.09	110.00	24.00	35.67	8.96
2009	59.62	10.31	93.17	13.54	109.00	24.00	34.00	8.16
2010	58.78	8.47	91.08	14.12	108.00	23.00	34.42	6.95
2011	58.37	9.75	89.17	12.30	106.00	19.00	34.00	8.55
2012	60.10	10.00	94.17	13.47	111.00	18.00	33.75	9.55
2013	60.27	10.66	92.83	12.91	110.00	15.00	34.08	9.95
2014	62.31	8.45	94.08	11.86	106.00	25.00	36.83	8.96
2015	61.17	10.09	93.08	11.85	106.00	20.00	36.08	10.85
2016	61.08	8.97	91.58	13.82	109.00	22.00	35.33	8.13
2017	61.74	10.35	93.50	14.60	114.00	19.00	36.08	9.89
2018	60.93	9.45	91.75	12.14	108.00	21.00	34.92	9.93
2019	60.06	9.85	88.83	13.82	105.00	23.00	33.17	7.90
2020	60.92	10.43	94.83	14.96	115.00	21.00	35.08	9.87
2021	60.11	9.71	93.75	11.79	111.00	24.00	34.50	8.87
2022	60.94	11.28	93.00	14.74	112.00	23.00	36.00	9.90
2023	59.33	9.74	88.75	14.57	110.00	26.00	35.58	8.38
2024	61.27	10.46	91.58	15.05	112.00	26.00	36.83	9.25
2025	52.25	4.21	84.60	5.41	91.00	24.00	33.00	8.00

Exceedances

When the seasonal max or min reaches a new high or low, indicate this with a true/false.

Code

out <- monthly_weather %>%
group_by(season) %>%
arrange(date) %>%
mutate(
start_flag = year > 1977,
cummax = cummax_ignore_na(max, start_flag),
cummin = cummin_ignore_na(min, start_flag),
new_max = cummax == max,
new_min = cummin == min
) %>%
ungroup()

monthly_weather$new_max <- out$new_max
monthly_weather$new_min <- out$new_min

We can either set this to start from the beginning of the series, or start the “tracker” at a certain specific time. Since we will have three parts to our music and we want the exceedances to be part 3, we will start the tracker in 1977 (see Sonifcation Design page for reasoning behind choice of 1977).

Code

out_all <- monthly_weather %>%
arrange(date) %>%
mutate(
start_flag = year > 1899,
cummax = cummax_ignore_na(max, start_flag),
cummin = cummin_ignore_na(min, start_flag),
new_max = cummax == max,
new_min = cummin == min
) %>%
ungroup()

monthly_weather$new_max_all_time <- out_all$new_max
monthly_weather$new_min_all_time <- out_all$new_min

monthly_weather %>%
ggplot(aes(x = date)) +
  geom_line(aes(y = min, color = "min")) +
  geom_point(aes(y = min, color = "min"), alpha = 0.1) +
  geom_line(aes(y = max, color = "max")) +
  geom_point(aes(y = max, color = "max"), alpha = 0.1) +
  geom_point(data = filter(monthly_weather, new_max_all_time),
             aes(x = date, y = max, color = "max"), size = 6, shape = 17) +
  geom_point(data = filter(monthly_weather, new_min_all_time),
             aes(x = date, y = min, color = "min"), size = 6, shape = 15) +
  scale_color_manual(values = plot_cols, name = "Exceedance\nType") + 
  scale_x_date(labels = monthly_weather$season_year[monthly_weather$season == "Winter" & monthly_weather$season_year %%5 ==0], 
               breaks = monthly_weather$xmin[monthly_weather$season == "Winter"& monthly_weather$season_year %%5 ==0]) +
  theme(axis.text.x = element_text(angle = 25))

Code

out <- monthly_weather %>%
arrange(date) %>%
mutate(
start_flag = year > 1977,
cummax = cummax_ignore_na(max, start_flag),
cummin = cummin_ignore_na(min, start_flag),
new_max = cummax == max,
new_min = cummin == min
) %>%
ungroup()

monthly_weather$new_max <- out$new_max
monthly_weather$new_min <- out$new_min

monthly_weather %>%
ggplot(aes(x = date)) +
  geom_line(aes(y = min, color = "min")) +
  geom_point(aes(y = min, color = "min"), alpha = 0.1) +
  geom_line(aes(y = max, color = "max")) +
  geom_point(aes(y = max, color = "max"), alpha = 0.1) +
  geom_point(data = filter(monthly_weather, new_max),
             aes(x = date, y = max, color = "max"), size = 6, shape = 17) +
  geom_point(data = filter(monthly_weather, new_min),
             aes(x = date, y = min, color = "min"), size = 6, shape = 15) +
  scale_color_manual(values = plot_cols, name = "Exceedance\nType") + 
  scale_x_date(labels = monthly_weather$season_year[monthly_weather$season == "Winter" & monthly_weather$season_year %%5 ==0], 
               breaks = monthly_weather$xmin[monthly_weather$season == "Winter"& monthly_weather$season_year %%5 ==0]) +
  theme(axis.text.x = element_text(angle = 25))

Most of the all time highs occur in summer, and most all-time lows occur in winter. This is unsurprising, and basically a consequence of the statistics we are looking at and the properties of the maximum and the minimum.

Seasonal Exceedances

Since our goal is to summarize data for sonification, we want to make sure the information does not get frontloaded all at the beginning. So, looking at seasonal exceedances gives us more chances to hear extremes.

Note that the unusually high fall temperatures are both for all time and if we start the “new extreme” timer in 1977.

All time seasonal exceedance
Seasonal exceedance since 1977

Code

out_all <- monthly_weather %>%
group_by(season)%>%
arrange(date) %>%
mutate(
start_flag = year > 1899,
cummax = cummax_ignore_na(max, start_flag),
cummin = cummin_ignore_na(min, start_flag),
new_max = cummax == max,
new_min = cummin == min
) %>%
ungroup()

monthly_weather$new_max_all_time <- out_all$new_max
monthly_weather$new_min_all_time <- out_all$new_min

monthly_weather %>%
ggplot(aes(x = date)) +
  geom_line(aes(y = min, color = "min")) +
  geom_point(aes(y = min, color = "min"), alpha = 0.1) +
  geom_line(aes(y = max, color = "max")) +
  geom_point(aes(y = max, color = "max"), alpha = 0.1) +
  geom_point(data = filter(monthly_weather, new_max_all_time),
             aes(x = date, y = max, color = season), size = 6, shape = 17) +
  geom_point(data = filter(monthly_weather, new_min_all_time),
             aes(x = date, y = min, color = season), size = 6, shape = 15) +
  scale_color_manual(values = plot_cols, name = "Exceedance\nType") + 
  scale_x_date(labels = monthly_weather$season_year[monthly_weather$season == "Winter" & monthly_weather$season_year %%5 ==0], 
               breaks = monthly_weather$xmin[monthly_weather$season == "Winter"& monthly_weather$season_year %%5 ==0]) +
  theme(axis.text.x = element_text(angle = 25))

Code

out <- monthly_weather %>%
group_by(season)%>%
arrange(date) %>%
mutate(
start_flag = year > 1977,
cummax = cummax_ignore_na(max, start_flag),
cummin = cummin_ignore_na(min, start_flag),
new_max = cummax == max,
new_min = cummin == min
) %>%
ungroup()

monthly_weather$new_max <- out$new_max
monthly_weather$new_min <- out$new_min

monthly_weather %>%
ggplot(aes(x = date)) +
  geom_line(aes(y = min, color = "min")) +
  geom_point(aes(y = min, color = "min"), alpha = 0.1) +
  geom_line(aes(y = max, color = "max")) +
  geom_point(aes(y = max, color = "max"), alpha = 0.1) +
  geom_point(data = filter(monthly_weather, new_max),
             aes(x = date, y = max, color = season), size = 6, shape = 17) +
  geom_point(data = filter(monthly_weather, new_min),
             aes(x = date, y = min, color = season), size = 6, shape = 15) +
  scale_color_manual(values = plot_cols, name = "Exceedance\nType") + 
  scale_x_date(labels = monthly_weather$season_year[monthly_weather$season == "Winter" & monthly_weather$season_year %%5 ==0], 
               breaks = monthly_weather$xmin[monthly_weather$season == "Winter"& monthly_weather$season_year %%5 ==0]) +
  theme(axis.text.x = element_text(angle = 25))

Missing Data

As you will hear (and see) in the animation, there are occasionally missing data points, resulting in silence!

Monthly level missing data

Code

## code inspired by https://r-graph-gallery.com/79-levelplot-with-ggplot2.html
x <- monthly_weather$date
y <- c("min", "avg", "max")
data <- expand.grid(X=x, Y=y)
data$Z <- c(ifelse(is.na(monthly_weather$min),"missing", "observed"),
            ifelse(is.na(monthly_weather$mean),"missing", "observed"),
            ifelse(is.na(monthly_weather$max),"missing", "observed"))
n_year <- nrow(monthly_weather)/12

x_breaks <- seq(from = ymd(monthly_weather$date[1]),
                to = ymd(monthly_weather$date[1392]), 
                length.out = floor(n_year/10))
x_labs <- year(x_breaks)

change_date_part_1 <- ymd("1910-01-01")
change_date_part_2 <- monthly_weather[which.max(monthly_weather$max),"date"] %>% pull()
change_date_part_3 <- ymd("1977-01-01")
  
data %>% tibble() %>%
mutate(X = as.Date(X, "%Y-%m-%d"))%>%
  ggplot(aes(x= X, y=factor(Y), fill = Z)) +
  geom_tile(height = 0.5) +
  scale_x_date(breaks = x_breaks, labels = x_labs) +
  scale_fill_manual(name = "",
    values = c("missing" = "black", "observed" = "grey90")
  ) + ylab("Summary \nStatistic") + xlab("Year (one line per month)") +
 theme(panel.grid.major = element_blank(), 
       panel.grid.minor = element_blank(),
       panel.background = element_blank(), 
       axis.line = element_blank(), 
       axis.title.y = element_text(angle = 0, vjust = .5))

Missing data hidden by aggregation

Since we are looking at monthly level data, presumably, each data point represents the summary (average, max, min) of 28, 29, 30, or 31 data points. However, there are exceptions

May 2025 values of all 3 summary statistics only include at most 9 points, since the data set was sourced May 9, 2025
- May 9 is based on the date of a raw pdf file I believe I saved using my browser’s print to pdf function
the NOWData tool only reports the monthly values, and does not indicate whether any of the daily (or lower level data that the daily values may be aggregated from) values were missing.
- So it’s possible some other (besides known May 2025) monthly data points represent fewer values than expected due to missing data. There may be documentation on the NOAA site about this, please let me know if you are aware!

Data for sonification

While busy, this is a visualization of the entire data set we will be working with for our sonification.

The full monthly history of the series is too much for one plot– this is why we want to animate and/or sonify it– to add a time dimension to our perceptualization and allowing us to “zoom in” as we did on the beginning of the series above.

Code

ggplot(monthly_weather, aes(x = date)) +
  geom_line(aes(y = min, color = "min")) +
  geom_point(aes(y = min, color = season)) +
  geom_line(aes(y = max, color = "max")) +
  geom_point(aes(y = max, color = season)) +
  geom_line(aes(y = mean, color = "mean")) +
  geom_point(aes(y = mean, color = season)) +
  geom_point(data = filter(monthly_weather, new_max),
             aes(x = date, y = max, color = season), size = 6, shape = 17) +
  geom_point(data = filter(monthly_weather, new_min),
             aes(x = date, y = min, color = season), size = 6, shape = 15) +
  scale_color_manual(values = plot_cols) + 
  scale_x_date(labels = monthly_weather$season_year[monthly_weather$season == "Winter" & monthly_weather$season_year %%5 ==0], 
               breaks = monthly_weather$xmin[monthly_weather$season == "Winter"& monthly_weather$season_year %%5 ==0]) +
  theme(axis.text.x = element_text(angle = 25))

Output data

This file is read in at the top of the sonification design details.

save(monthly_weather, file = "data/monthly_weather")
save(plot_cols, file = "data/plot_cols")