这里是另一个选项。ggplot有一个geom_point，其中数据过滤到Magnitude > 5。这在aes中有color = Magnitude。这使得可以使用scale_color_gradiant2设置这些点的颜色样式，如下所示。其他数据也会被过滤（Magnitude <= 5）和Magnitude被重命名为mag。这样就可以独立于其他点设置这些点的样式。您可以使用geom_point。但这可能会导致一些重叠，并给人以alpha未应用的印象。我使用geom_jitter来避免过度绘制。
我也采取了自由使用不同的大小为不同的子集和一些颜色。

library(tidyverse)

ggplot() +
  geom_point(data = data |> filter(Magnitude > 5), aes(
    Date, Magnitude,
    color = Magnitude 
  ), size = 3, show.legend = FALSE) +
  # here you can also use geom_point
  geom_jitter(data = data |>
    filter(Magnitude <= 5) |> rename(mag = Magnitude), aes(
    Date, mag,
    alpha = mag
  ), color = "grey", size = 1, show.legend = FALSE) +
  scale_colour_gradient2(
    low = "green",
    mid = "orange",
    high = "red",
    midpoint = 6,
    guide = "none",
    aesthetics = "colour"
  )

你可以用分段来伪造你的alpha。这比使用scale_alpha（在我看来）给你更好的控制。
我已经调整了外观，使其更接近你的链接示例图。诀窍是将alpha值分配给你的幅值。在这样做的数据框中，你可以影响alpha值在哪里会有什么值。

library(ggplot2)
df <- data.frame(Date = rep(seq(as.Date("2021-01-01"), as.Date("2021-12-31"), by = "day"), 10),
                 Magnitude = runif(3650, max = 5))
## generate your alpha segments
df_seg <- data.frame(y = seq(min(df$Magnitude), max(df$Magnitude), len = 1000), 
                     alpha = seq(1, 0.3, len = 1000))

ggplot(df, aes(Date, Magnitude)) +
  ## use alpha as aesthetic
  ## color for the same look as in your linked example graph
  geom_point(aes(color = Magnitude),
    size = 1.7, shape = 16, stroke = 0
  ) +
  geom_segment(
    data = df_seg, aes(
      y = y, yend = y,
      x = as.Date("2021-01-01"),
      xend = as.Date("2021-12-31"),
      alpha = alpha
    ),
    ## you need to play around with the line width a bit
    linewidth = .2,
    color = "white", 
    ## to remove the alpha legend
    show.legend = F
  ) +

  ## add some random divergent colors
  scale_color_gradientn(values = c(0, .4, 1), colors = c("darkgreen", "darkgreen", "red")) +
  theme_classic() +
  coord_cartesian(expand = F)

我是说我发现添加段来伪造alpha更容易。正如用户AllanCameron正确指出的，这是有争议的。直接将alpha添加到scale_color_gradientn中的颜色也会给予人一个很好的视觉效果，而且确实要短得多。

ggplot(df, aes(Date, Magnitude)) +
  geom_point(aes(color = Magnitude),
             size = 1.7, shape = 16, stroke = 0
  ) +
  scale_color_gradientn(values = c(0, .4, 1), 
                        colors =c(alpha("darkgreen", 0.1), alpha("darkgreen", 0.5), "red")) +
  theme_classic() +
  coord_cartesian(expand = F)

可以通过更改scale_alpha()中的low和high参数来调整透明度渐变。要使震级较低的地震具有较浅的色调，可以增加low值，例如：

ggplot(df,aes(Date, Magnitude)) +
  geom_point(data = df %>% filter(Magnitude > 5),
             alpha = 0.9, size = 1.7, shape = 16, stroke = 0,
             color = "#264653")+
  geom_point(aes(color = Magnitude),
             data = df %>% filter(Magnitude <= 5),
             alpha = 0.5, size = 1.7, shape = 16, stroke = 0) +
  scale_alpha(limits = c(0, 5), range = c(0.1, 0.5))