-
Notifications
You must be signed in to change notification settings - Fork 2
/
dune-analysis.R
252 lines (199 loc) · 11.9 KB
/
dune-analysis.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
library(ggplot2)
library(gridExtra)
library(nlme)
library(mgcv)
library(data.table)
library(dplyr)
library(grid)
#library(moments)
#library(foreign)
source("dune-analysis_funs.R")
#dune_data <- read.dbf("transects_dunes_points_dtm_slope_dens.dbf", as.is=T)
#save(dune_data, file="dune_data.RData")
load("dune_data.RData")
shrub_ts_data <- read.csv("historical_shrub_data.csv")
# add transect length/height -----------------------------------------------------
dune_data_seq <- dune_data %>%
filter(h_asl > 0) %>%
filter(!dune %in% c("dune5", "dune7a", "dune2"),
!Id %in% c(157,158,159,161)) %>% # few odd length transects
rename(shrub_dens = RASTERVALU) %>%
group_by(Id) %>% #transect number
mutate(distance = seq(0, by=0.1, length.out = n()),
height = sapply(h_asl, function(x) x - min(h_asl)),
r_j = height - mean(height),
q_j = slope - mean(slope)) %>%
ungroup() %>%
filter(height > 0.20,
!(distance > 175 & dune == "dune2")) # lengths too uneven
# dune 9/7 transects were switched somehow
#dune_data_seq$distance[dune_data_seq$dune == "dune7a" & dune_data_seq$fence == "inside"] <- abs(dune_data_seq$distance[dune_data_seq$dune == "dune7a" & dune_data_seq$fence == "inside"] - max(dune_data_seq$distance[dune_data_seq$dune == "dune7a" & dune_data_seq$fence == "inside"]))
dune_data_seq$distance[dune_data_seq$dune == "dune7b" & dune_data_seq$fence == "outside"] <- abs(dune_data_seq$distance[dune_data_seq$dune == "dune7b" & dune_data_seq$fence == "outside"] - max(dune_data_seq$distance[dune_data_seq$dune == "dune7b" & dune_data_seq$fence == "outside"]))
dune_data_seq$distance[dune_data_seq$dune == "dune9" & dune_data_seq$fence == "outside"] <- abs(dune_data_seq$distance[dune_data_seq$dune == "dune9" & dune_data_seq$fence == "outside"] - max(dune_data_seq$distance[dune_data_seq$dune == "dune9" & dune_data_seq$fence == "outside"]))
# calculate metrics -------------------------------------------------------
dune_shrub_metrics <- dune_data_seq %>%
group_by(dune, fence, Id) %>%
summarise(shrub_dens = mean(shrub_dens) * 10000, # transform to shrubs/hectare
dune_area = sum(height) * 0.1,
dune_width = n() * 0.1,
height_max = max(height),
height_kurt = sum(r_j^4) / (n() * (sqrt(mean(r_j^2)))^4),
heightdev_rms = sqrt(mean(r_j^2)),
slope_rms = sqrt(mean(q_j^2)),
slope_cv = sd(slope) / mean(slope),
flatness = log(dune_area / height_max),
flatness2 = dune_width / height_max,
sphericity = dune_area / (pi*height_max*height_max/2),
rectangularity = dune_area / (dune_width * height_max),
east = median(east), north = median(north)) %>%
filter(height_max > 2) # any odd little transects
# as factor for modelling
dune_shrub_metrics$dune <- factor(dune_shrub_metrics$dune) # need factors for some mgcv args
dune_shrub_metrics$fence <- factor(dune_shrub_metrics$fence)
# secondary data set with more balanced shrub density insdie/outside
dune_shrub_metrics2 <- dune_shrub_metrics %>%
filter(shrub_dens < max(dune_shrub_metrics$shrub_dens[dune_shrub_metrics$fence=="outside"]))
# dune transect plots -----------------------------------------------------
transect_plot <- ggplot(data = dune_data_seq, aes(x = distance, y = height, group = Id)) +
geom_line(aes(colour = fence), size = 1, alpha = 0.2) +
scale_color_manual(values = c("red", "blue")) +
ylab("Dune height (m)") + xlab("Distance along transect (m)") + theme_bw() +
facet_grid(dune ~ ., scales = "fixed")
ggsave("plots/dune_transect_plot.png", transect_plot, width = 30, height = 20, units = "cm")
# shrub time series -------------------------------------------------------
shrub_timeseries <- shrub_ts_data %>%
mutate(#year = factor(year),
shrub_density = count / (area/10000)) %>% # shrubs/hectare
group_by(fence, year) %>%
summarise(mean_shrub_density = mean(shrub_density),
std_dev = sd(shrub_density),
nsamps = n(),
std_err = std_dev / sqrt(nsamps))
shrub_diff_ts <- rbindlist(lapply(X = unique(shrub_timeseries$year), FUN = shrub_dens_diff, shrub_timeseries))
ggplot(shrub_timeseries, aes(x = year, y = mean_shrub_density, colour = fence)) +
geom_line(aes(group = fence), size = 1) +
geom_errorbar(aes(ymax = mean_shrub_density + std_err, ymin = mean_shrub_density - std_err), size = 0.25) +
scale_x_continuous(breaks = unique(shrub_timeseries$year)) +
theme_classic()
history_plt <- ggplot(shrub_diff_ts, aes(x = year, y = diff)) +
geom_line(size = 1) +
geom_errorbar(aes(ymax = diff + std_err, ymin = diff - std_err), size = 0.25, width = 0.4) +
scale_x_continuous(breaks = unique(shrub_diff_ts$year)) +
ylim(-10, 80) +
geom_hline(yintercept = 0, linetype = 2) +
ylab("Shrub density (shrubs/Ha)") + xlab("Image date") +
#ggtitle("Difference in shrub density inside and outside the Dingo fence") +
# annotate("text", x = 1948, y = 8,
# label = "Difference in shrub density inside and outside the Dingo fence", size = 6, hjust = 0) +
#annotate("text", x = 1992, y = -5, label = sprintf('\u2193'), size = 9) +
geom_segment(aes(x = 1992, y = -2, xend = 1992, yend = -8), arrow = arrow(length = unit(0.03, "npc"))) +
annotate("text", x = 1994, y = -5, label = "more shrubs outside", size = 4, hjust = 0) +
#annotate("text", x = 1992, y = 5, label = sprintf('\u2191'), size = 9) +
geom_segment(aes(x = 1992, y = 2, xend = 1992, yend = 8), arrow = arrow(length = unit(0.03, "npc"))) +
annotate("text", x = 1994, y = 5, label = "more shrubs inside", size = 4, hjust = 0) +
theme_classic()
ggsave(filename = "plots/history_plot.pdf", device = "pdf", plot = history_plt, width = 9, height = 6)
# shrub metric plots ------------------------------------------------------
boxplot(shrub_dens ~ fence, data = dune_shrub_metrics, ylab = "Shrub density")
my_violin("shrub_dens", dune_shrub_metrics, ylab = "Shrub density")
shrub_plt <- my_hist("shrub_dens", dune_shrub_metrics, ylab = "Shrub density", legend_off = F)
ggsave("plots/shrubiness.png", device = "png", plot = shrub_plt, width = 6, height = 4)
shrub_model <- lme(fixed = shrub_dens ~ fence, random = ~ 1 | dune,
correlation = corGaus(form = ~ east + north | dune, nugget = F),
data = dune_shrub_metrics)
summary(shrub_model)
par(mfrow=c(3,2))
boxplot(height_max ~ fence, data = dune_shrub_metrics, ylab = "Maximum height")
boxplot(heightdev_rms ~ fence, data = dune_shrub_metrics, ylab = "Height dev.")
boxplot(slope_rms ~ fence, data = dune_shrub_metrics, ylab = "Slope-iness")
boxplot(slope_cv ~ fence, data = dune_shrub_metrics, ylab = "Slope CV")
boxplot(flatness ~ fence, data = dune_shrub_metrics, ylab = "Flatness")
boxplot(rectangularity ~ fence, data = dune_shrub_metrics, ylab = "Rectangularity")
metric_plt <- grid.arrange(
my_violin("height_max", data = dune_shrub_metrics, ylab = "Maximum height"),
my_violin("heightdev_rms", data = dune_shrub_metrics, ylab = "Deviation in height profile"),
my_violin("slope_rms", data = dune_shrub_metrics, ylab = "Absolute range in slope"),
my_violin("slope_cv", data = dune_shrub_metrics, ylab = "Slope variability"),
my_violin("flatness", data = dune_shrub_metrics, ylab = "Flatness"),
my_violin("rectangularity", data = dune_shrub_metrics, ylab = "Rectangularity"),
ncol = 3)
ggsave("plots/dune_metric_plots.png", device = "png", plot = metric_plt, width = 12, height = 8)
metric_dens_plt <- grid.arrange(
my_hist("height_max", data = dune_shrub_metrics, ylab = "Maximum height"),
my_hist("heightdev_rms", data = dune_shrub_metrics, ylab = "Deviation in height profile"),
my_hist("slope_rms", data = dune_shrub_metrics, ylab = "Absolute range in slope"),
my_hist("slope_cv", data = dune_shrub_metrics, ylab = "Slope variability", legend_off = F),
my_hist("flatness", data = dune_shrub_metrics, ylab = "Flatness"),
my_hist("rectangularity", data = dune_shrub_metrics, ylab = "Rectangularity"),
ncol = 3)
ggsave("plots/dune_metric_dens-plots.pdf", device = "pdf", plot = metric_dens_plt, width = 12, height = 8)
# geomorph clustering -----------------------------------------------------
library(ggbiplot)
pca_dat <- dune_shrub_metrics %>%
ungroup() %>%
select(height_max, heightdev_rms, slope_rms, slope_cv, flatness, rectangularity)
pca <- prcomp(pca_dat, scale. = T)
# pca_plot_data <- data.frame(pca$x, shrub_density = dune_shrub_metrics$shrub_dens,
# dune = dune_shrub_metrics$dune, fence = dune_shrub_metrics$fence)
# ggplot(data = pca_plot_data, mapping = aes(x = PC1, y = PC2, col = fence)) +
# geom_point(size = 2, alpha = 0.5) +
# theme_classic()
pca_plt <- ggbiplot(pca, obs.scale = 1, var.scale = 1,
groups = dune_shrub_metrics$dune,
ellipse = F, circle = F, alpha = 0) +
geom_point(aes(shape = dune_shrub_metrics$fence, colour = dune_shrub_metrics$dune), size = 3) +
scale_color_brewer(palette = "Set1", name = 'Dune number') +
scale_shape_discrete(name = 'Dingo fence') +
theme_classic()
ggsave("plots/dune_metric_pca.png", device = "png", plot = pca_plt, width = 12, height = 10)
ggbiplot(pca, obs.scale = 1, var.scale = 1,
groups = dune_shrub_metrics$shrub_dens,
ellipse = F, circle = F, alpha = 0) +
geom_point(aes(colour = dune_shrub_metrics$shrub_dens, shape = dune_shrub_metrics$fence), size = 3) +
scale_color_gradient(low = "coral", high = "green", name = "Shrub density") +
scale_shape_discrete(name = 'Dingo fence') +
theme_classic()
# dune and shrub models ---------------------------------------------------
# shrubiness across fence - YES
dune_shrub_metrics$shrub_dens_log <- log1p(dune_shrub_metrics$shrub_dens)
shrubiness <- gamm(formula = shrub_dens ~ fence + s(dune, bs = 're'),
correlation = corGaus(form = ~ east + north, nugget = F),
data = dune_shrub_metrics, method = "REML")
summary(shrubiness$gam)
plot(shrubiness$lme, all.terms = T)
plot(predict(shrubiness$gam, type="response") ~ dune_shrub_metrics$fence)
# shrubiness across space - YES
spatial_smooth <- gam(formula = shrub_dens ~ te(east,north), data = dune_shrub_metrics)
summary(spatial_smooth); plot(spatial_smooth, residuals = T)
## models
# make_plots <- F
# print_summary <- F
height <- shrub_smooth_dune_re("height_max", dune_shrub_metrics, bs = "'tp'")
heightdev <- shrub_smooth_dune_re("heightdev_rms", dune_shrub_metrics, bs = "'tp'")
sloperms <- shrub_smooth_dune_re("slope_rms", dune_shrub_metrics, bs = "'tp'")
slopecv <- shrub_smooth_dune_re("slope_cv", dune_shrub_metrics, bs = "'tp'")
flat <- shrub_smooth_dune_re("flatness", dune_shrub_metrics, bs = "'tp'")
rect <- shrub_smooth_dune_re("rectangularity", dune_shrub_metrics, bs = "'tp'")
my.gam.check(heightdev)
my.gam.check(height)
my.gam.check(sloperms)
my.gam.check(slopecv)
my.gam.check(flat)
my.gam.check(rect)
get_paper_values(height)
get_paper_values(heightdev)
get_paper_values(sloperms)
get_paper_values(slopecv)
get_paper_values(flat)
get_paper_values(rect)
# parial plots
heightplt <- plot_shrub_by_fence(height, dune_shrub_metrics, ylabel = "Maximum height", panel_lab = "a.")
heightdevplt <- plot_shrub_by_fence(heightdev, dune_shrub_metrics, ylabel = "Height deviation", panel_lab = "b.")
slopermsplt <- plot_shrub_by_fence(sloperms, dune_shrub_metrics, ylabel = "Roughness (slope rms)", panel_lab = "x.")
slopecvplt <- plot_shrub_by_fence(slopecv, dune_shrub_metrics, ylabel = "Roughness (slope CV)", panel_lab = "c.")
flatplt <- plot_shrub_by_fence(flat, dune_shrub_metrics, ylabel = "Flatness", panel_lab = "y.")
rectplt <- plot_shrub_by_fence(rect, dune_shrub_metrics, ylabel = "Rectangularity", panel_lab = "d.")
shrub_partials_all <- grid.arrange(heightplt, heightdevplt, slopermsplt, slopecvplt, flatplt, rectplt)
shrub_partials <- grid.arrange(heightplt, heightdevplt, slopecvplt, rectplt)
ggsave("plots/shrub_partials_all.png", device = "png", plot = shrub_partials_all, width = 12, height = 10)
ggsave("plots/shrub_partials.pdf", device = "pdf", plot = shrub_partials, width = 12, height = 7)