ggplot(fire, aes(x = GeneralCause, y = EstTotalAcres)) +
geom_bar(stat = "identity", fill = "darkred") +
labs(x = "General Cause of Fire",
y = "Estimated Total Acres",
title = "Estimated Total Acres by General Cause of Fire") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))+
scale_y_continuous(labels = scales::comma, breaks = scales::pretty_breaks(n = 10))
## Warning: Removed 79 rows containing missing values (`position_stack()`).
In the barchart depicting fire causes by acreage burned, it’s evident that lightning stands out as the primary contributer to fires. I wanted to learn more about how lightning was the main cause of fire throughout the years.
palette <- brewer.pal(n = length(unique(fire$GeneralCause)), name = "Paired")
ggplot(fire, aes(x = FireYear, y = EstTotalAcres, fill = GeneralCause)) +
geom_bar(stat = "identity") +
labs(x = "Fire Year",
y = "Estimated Total Acres",
title = "Estimated Total Acres by Fire Year and Fire Cause",
fill = "Fire Cause") +
scale_fill_manual(values = palette) +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
scale_y_continuous(labels = scales::comma, breaks = scales::pretty_breaks(n = 10))
## Warning: Removed 79 rows containing missing values (`position_stack()`).
From this graph it was evident that fire was the main cause for lightning up until 2020, where the cause of fire is under investigation.
selected_columns <- dplyr::select(fire, FireYear, GeneralCause, EstTotalAcres)
top_cause_table <- selected_columns %>%
group_by(FireYear) %>%
top_n(1, EstTotalAcres) %>%
arrange(FireYear)
print(top_cause_table)
## # A tibble: 24 × 3
## # Groups: FireYear [23]
## FireYear GeneralCause EstTotalAcres
## <int> <chr> <dbl>
## 1 2000 Lightning 56319
## 2 2001 Lightning 32352
## 3 2002 Lightning 499945
## 4 2002 Lightning 499945
## 5 2003 Lightning 90376
## 6 2004 Equipment Use 4705
## 7 2005 Lightning 42700
## 8 2006 Lightning 53523.
## 9 2007 Lightning 140360
## 10 2008 Lightning 4891.
## # ℹ 14 more rows
I wanted to create a table to give me the exact numbers of the top cause for fire throughout the years.
ggplot() +
geom_sf(data = fireshp, fill = "red", color = "black") +
geom_sf(data = or_shape, fill = NA, color = "black", size = 1) +
labs(title = "Fire Distribution in Oregon") +
theme_grey()
This graph displays the total distribution of Oregon Fires from 2000-2022.
lightning_fires <- fireshp[fireshp$humanorlig == "Lightning", ]
human_fires <- fireshp[fireshp$humanorlig == "Human", ]
ggplot() +
geom_sf(data = human_fires, aes(fill = "Human"), color = "orange", alpha = 0.5) +
geom_sf(data = lightning_fires, aes(fill = "Lightning"), color = "blue", alpha = 0.5) +
geom_sf(data = orsf, fill = "NA", color = "black", size = 1) +
scale_fill_manual(values = c("orange", "blue"),
name = "Cause of Fire",
labels = c("Human", "Lightning")) +
labs(title = "Human vs Lightning-Caused Fires in Oregon") +
theme_grey()
From the previous bar charts I wanted to visualize the distribution of lightning vs. human caused fires in Oregon.
largest_fire <- fire %>%
top_n(1, EstTotalAcres)
largest_fire <- fireshp %>%
top_n(1, esttotalac)
print(largest_fire)
## Simple feature collection with 2 features and 43 fields
## Geometry type: POINT
## Dimension: XY
## Bounding box: xmin: -123.8919 ymin: 42.40541 xmax: -123.8919 ymax: 42.40541
## Geodetic CRS: WGS84(DD)
## serial firecatego fireyear area districtna unitname fullfirenu
## 1 64792 STAT 2002 SOA Southwest Oregon Grants Pass 02-712061-03
## 2 65367 STAT 2002 SOA Coos - CFPA Gold Beach 02-723024-03
## firename size_class esttotalac protected_ humanorlig causeby
## 1 ODF / BISCUIT G 499945 10226.00 Lightning Lightning
## 2 Biscuit Private G 499945 273.94 Lightning Lightning
## generalcau specificca cause_comm lat_dd long_dd fo_landown twn rng sec
## 1 Lightning Lightning <NA> 42.40541 -123.8919 USFS 36S 10W 28
## 2 Lightning Lightning <NA> 42.40541 -123.8919 USFS 36S 10W 28
## subdiv landmarklo county regusezone reguserest
## 1 SWSW 19 NW OF CAVE JUNCTION Josephine SW3 Reg Use Closure
## 2 SWSW 1 S Agness Josephine SW3 Reg Use Closure
## industrial date_ign_d time_ign_d date_repor time_repor
## 1 Lvl 1 Fire Season Only 2002-07-13 14:30:00.000 2002-07-15 12:14:00.000
## 2 Lvl 1 Fire Season Only 2002-07-13 14:30:00.000 2002-07-15 12:14:00.000
## date_disco time_disco date_contr time_contr date_creat time_creat
## 1 2002-07-15 12:13:00.000 2002-11-08 18:00:00.000 2002-08-19 18:28:00.000
## 2 2002-07-15 12:13:00.000 2002-11-08 18:00:00.000 2002-09-16 19:46:00.000
## date_modif time_modif districtco unitcode distfirenu
## 1 2002-12-09 16:46:00.000 71 712 061
## 2 2003-01-30 11:18:00.000 72 723 024
## geometry
## 1 POINT (-123.8919 42.40541)
## 2 POINT (-123.8919 42.40541)
This table shows that the Biscuit Fire was the largest fire in Oregon in 2020.
fire_data_20204<- subset(fireshp, fireyear == 2020)
ggplot() +
geom_sf(data = orsf, fill = "lightgrey") +
geom_sf(data = fireshp, color = "black", size = 0.5) +
geom_sf(data = fire_data_20204, color = "red", size = .5) +
labs(title = "Fires in Oregon (Year 2020 Highlighted in Red)", x = "Longitude", y = "Latitude") +
scale_fill_manual(name = "Data Source",
values = c("Fire Shapefile" = "black", "Fire Data 2020" = "red")) +
theme_minimal()
I chose to display the total fires in 2020 vs the total fires in Oregon because the most amount of firs in Oregon were seen in 2020. The spatial distribution reveals that the fires were spread out across all of Oregon within the distribution pattern.
##Discussion
When looking at the data I wanted to understand the cause of fire and their associated acreage. After plotting a graph of fire causes and total fire acreage I found that lightning was the main cause seen across 22 year in Oregon. I then used a bar graph to identify the different distributions of fire causes during the time period. I then used a table to produce quantitative data of largest fire cause and acreage of each year in Oregon.
The second section of my project was dedicted to spatially visualizing the data I was using. I first plotted a map of Oregon with all of the data to understand the spatial distribution. I found that there was either a lack of fire in the south easter part of Oregon or lack of data becauase there were a lack of points in this region. Because there was existing data on human vs lightning caused fires in the dataset I wanted to plot this to see the differences. I found that lightning had more fire causes than human. I also wanted to identify the largest fire in 2020 which was caused by lightning.My last map displayed the total fires from 2020 in red compared to the total fires from the rest of the years in black because the majority of fires were from 2020.
Together these different visualizations emphasize the pressing need for proactive measures in fire management and prevention to mitigate the neviornmentl impacts of wildfires in Oregon. I was a bit shocked to learn that lightning was the main cause of fire, and while that may only be because of the dataset it is all the more reason to be concerned about future environmental hazards.
##Sources:
My data was titled ODF Fire Occurence Data 2000-2022 and was from Oregon.GOv Open Data Portal.I exported the data as a shapefile and csv. I also got my oregon shapefile from the US Census Bureau. The Oregon Fire Dataset contained data on 38 different fields that ranged from fire year to estimated total fire acres. Because the data was only displayed in points I could not display it in polygon format as I had previously wanted.
The website was titled : https://data.oregon.gov/Natural-Resources/ODF-Fire-Occurrence-Data-2000-2022/fbwv-q84y/about_data
Note that the echo = FALSE
parameter was added to the
code chunk to prevent printing of the R code that generated the
plot.