Health impacts of fine particles in Rauma

This page is a knowledge crystal of subtype variable. The page identifier is Op_en5841
Moderator:Jouni (see all)

Upload data {{#opasnet_base_link:Op_en5841}}

Question

How large are the emissions and health impacts of fine particles in Rauma, Finland? The focus is on emissions from the port of Rauma.

Answer

A recent model run

**Premature deaths per year**
Urban planning option	Result
Option 1 (Centre and nearby areas	2.83
Option 2 (From centre to sea)	2.28
Option 3 (Shores and paths)	2.79

More precise descriptions of the urban planning options that are currently under evaluation can be found from op_fi:Rauman sataman laajennuksen vaikutus terveyteen an Opasnet page (in Finnish only).

Rationale

⇤--#:: . The distribution model used has a resolution of 1 km only. This is not precise enough. The model only gives rough estimates and probably underestimates impacts. --Jouni Tuomisto 25. October 2012 00.24 (EEST) (type: truth; paradigms: science: attack)

⇤--#:: . It is true that the model underestimates exposure from sources closer than a few hundred metres. However, the direction of the bias is known, and in this case it undersestimates the results of especially Option 1 and 3, which contain more traffic in the city centre. Even with the current precision, the impacts seem higher than in option 2. So, the model fulfils its purpose to answer the question about whether there are differences in health impacts between the three options. There seem to be. --Jouni Tuomisto 25. October 2012 11.26 (EEST) (type: truth; paradigms: science: attack)

Calculations

NOTE! Currently the model calculates fine particle concentrations and exposures only within 10-15 km from the source.

Key for the default model run: BJ9p0oAwwPRGJ2sg (1000 iterations, 4 min 20 s)

+ Show code - Hide code

library(OpasnetUtils)
library(ggplot2)
library(xtable)
library(OpasnetUtilsExt)
library(rgdal)
library(maptools)
library(RColorBrewer)
library(classInt)
library(raster)

# Graph settings?

par(mfrow=c(6,1), mar=c(3,1,0,1), cex=1.5)

colorstrip <- function(colors, labels) {
	count <- length(colors)
	m <- matrix(1:count, count, 1)
	image(m, col=colors, ylab="", axes=FALSE)
	axis(1,approx(c(0, 1), n=length(labels))$y, labels)
}

# suorite Traffic volumes

data <- tidy(op_baseGetData("opasnet_base", "Op_fi3190"), objname = "suorite")
data <- data[data$Ajoneuvo != "Laiva", ] # Laivoista on suoraan PM-päästöt, joten ei lasketa suoritteesta.
data$suoriteResult <- data$suoriteLiikennemäärä * as.numeric(as.character(data$suoritePituus))

suorite <- new("ovariable",
	name = "suorite",
	data = data
)

if(intermediates) {cat("Traffic volume.\n"); print(xtable(suorite@data), type = 'html')}

# Päästö Emissions

Päästö <- new(
	"ovariable", 
	name = "Päästö",
	dependencies = data.frame(Name = "tieliikennepäästöt", Key = "0194s0uuucjxq8Wi"),
	formula = function(dependencies, ...) {
		ComputeDependencies(dependencies, ...)

		out <- tieliikennepäästöt * 365 * 1E-6 # Muutetaan päiväpäästö vuosipäästöksi ja grammat tonneiksi

		return(out)
	}
)

# Muita tarpeellisia arvoja Other important values

bg.mort <- 45182 / 5203826 # same values as used in PILTTI

erf <- 0.0097 # J. T. Tuomisto, A. Wilson, et al. Uncertainty in mortality response to airborne fine particulate matter... 2008
# unit: m^3 /ug

# Ovariablet

## Pitoisuudet Concentrations

Pitoisuus <- new(
	"ovariable", 
	name = "Pitoisuus",
	dependencies = data.frame(
		Name = c("Päästö", "LO", "LA")
	),
	formula =  function(dependencies, ...) {
		ComputeDependencies(dependencies, ...)

		temp <- GIS.Concentration.matrix(Päästö, LO, LA, ...)

		return(temp)
	} 
)

## Altistuminen Exposure

Altistuminen <- new(
	"ovariable", 
	name = "Altistuminen", 
	dependencies = data.frame(
		Name = c("Pitoisuus", "LO", "LA")
	),
	formula = function(dependencies, ...) {
		ComputeDependencies(dependencies, ...)

		out <- GIS.Exposure(Pitoisuus, LO, LA, ...)

		return(out)
	}
)

## Terveysvaikutukset Health impacts

Terveysvaikutukset <- new(
	"ovariable", 
	name = "Terveysvaikutukset", 
	dependencies = data.frame(
		Name = c("Altistuminen", "erf", "bg.mort")
	),
	formula = function(dependencies, ...) {
		ComputeDependencies(dependencies, ...)
	
		out <- Altistuminen * erf * bg.mort
	
		return(out)
	}
)

Päästö.temp <- EvalOutput(Päästö, N = N)
temp <- Päästö.temp@output[Päästö.temp@output$Saaste == "PM", ] # Valitaan vain PM-rivit.

if(intermediates) {cat("Detailed emission data (no ships).\n"); print(xtable(temp), type = 'html')}

temp <- as.data.frame(as.table(tapply(temp$PäästöResult, temp[c("LA", "LO", "Vaihtoehto", "Saaste", "PäästöSource")], sum)))
temp <- temp[!is.na(temp$Freq), ] # Pudotetaan tyhjät rivit pois.
colnames(temp)[colnames(temp) == "Freq"] <- "PäästöResult"

# Laivoista on suoraan PM-päästöt, joten ne otetaan tässä mukaan päästöihin.
# We have directly emission estimates from ships, so that data is taken in to the emissions.

laivat <- tidy(op_baseGetData("opasnet_base", "Op_fi3190"))
laivat <- laivat[laivat$Ajoneuvo == "Laiva", c("Vaihtoehto", "LA", "LO", "Liikennemäärä")] 

colnames(laivat)[colnames(laivat) == "Liikennemäärä"] <- "PäästöResult"
laivat$Saaste <- "PM"
laivat$PäästöSource <- "Data"

if(intermediates) {cat("Ship emissions.\n"); print(xtable(laivat), type = 'html')}

Päästö.temp@output <- rbind(temp, laivat)

cat("Total fine particle emissions from different locations.\n")
print(xtable(Päästö.temp@output), type = 'html')

Lopputulos <- data.frame() # Tähän taulukkoon kerätään data Table for end results.

for(i in 1:nrow(Päästö.temp@output)) {

	# Tässä käydään läpi yksi päästöpiste ja vaihtoehto kerrallaan ja lasketaan sille pitoisuus, altistus ja terveysvaikutus.
	# This loop goes through each emission location at a time and calculates concentrations, exposures, and health impacts.

	Päästö <- Päästö.temp@output$PäästöResult[i]
	LA <- as.numeric(as.character(Päästö.temp@output$LA[i]))
	LO <- as.numeric(as.character(Päästö.temp@output$LO[i]))

	# EvalOutput käyttää tilapäisiä, rivikohtaisia tietoja. Kunkin rivin tulos lisätään lopputulokseen.
	# EvalOutput uses temporary, row-specific information. The result of each row is added to the end result table.

	Lopputulos <- rbind(
		Lopputulos, 
		merge(
			Päästö.temp@output[i, ], 
			EvalOutput(Terveysvaikutukset, N = N)@output, by = NULL
		)
	)

	# Nollataan Altistuminen ja Pitoisuus, jotta malli laskee ne uusiksi uudelle päästölle.
	# The concentration and exposure ovariables must be emptied so that they are calculated again for a new emission.

	Altistuminen@output <- data.frame()
	Pitoisuus@output <- data.frame()

}

if(intermediates) {cat("Detailed end result.\n"); print(xtable(Lopputulos[Lopputulos$Iter == 1, ]), type = 'html')}

temp <- as.data.frame(as.table(tapply(Lopputulos$TerveysvaikutuksetResult, Lopputulos[c("Vaihtoehto", "Iter")], sum)))

if(N > 10) {ggplot(temp, aes(x = Freq, fill = Vaihtoehto)) + geom_density(alpha = 0.2)}

cat("Premature deaths in Rauma caused by fine particle emissions. (cases per year)\n")

print(xtable(as.data.frame(as.table(tapply(temp$Freq, temp[c("Vaihtoehto")], mean)))), type = 'html')

#################################

# Draw a concentration map.

cat("An example concentration field based on emissions from one location.\n")

temp <- EvalOutput(Pitoisuus, N = 1)

temp2 <- gsub("\\(", "", temp@output$LObin)
temp2 <- gsub("\\]", "", temp2)
temp2 <- strsplit(temp2, ",")

temp3 <- c()

for(i in 1:length(temp2)) {
	a <- as.numeric(temp2[[i]][1])
	b <- as.numeric(temp2[[i]][2])
	temp3[i] <- ((a + b) / 2)
}

temp@output$LO <- temp3

temp2 <- gsub("\\(", "", temp@output$LAbin)
temp2 <- gsub("\\]", "", temp2)

temp2 <- strsplit(temp2, ",")

temp3 <- c()

for(i in 1:length(temp2)) {
	a <- as.numeric(temp2[[i]][1])
	b <- as.numeric(temp2[[i]][2])
	temp3[i] <- ((a + b) / 2)
}

temp@output$LA <- temp3

#print(temp)

data <- data.frame(LO=temp@output$LO, LA=temp@output$LA, concentration=temp@output$PitoisuusResult)

data$truncated_concentration <- ifelse(data$concentration > 1, 1, data$concentration)

# Plot the data
coordinates(data)=c("LO","LA")
proj4string(data)<-("+init=epsg:4326")
epsg4326String <- CRS("+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs")
shp<-spTransform(data,epsg4326String)

#Create blank raster
rast<-raster()

#Set raster extent to that of point data
extent(rast)<-extent(shp)

#Choose number of columns and rows
ncol(rast) <- 41
nrow(rast) <- 41

#cat("<span style='font-size: 1.2em;font-weight:bold;'>PM2.5</span>\n")

#Rasterize point data
rast2<-rasterize(shp, rast, shp$truncated_concentration, fun=mean)

steps <- c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1)
colors <- rev(rainbow(length(steps), start=0, end=0.50))

colorstrip(colors, steps)

#Plot data
google.show_raster_on_maps(rast2, col=colors, style="height:500px;")

Ovariables (open assessment variables) used in the model:

Health impacts (Terveysvaikutukset): inputs: Altistuminen * erf * bg.mort
- Exposure-response function for PM (erf): constant, is given in the code
- Background mortality (bg.mort): constant, given in the code
- Exposure (Altistuminen): inputs: Pitoisuus, LO, LA
  - Uses function GIS.Exposure(Pitoisuus, LO, LA, ...)
  - Longitude (LO) and latitude (LA): given within the ovariable concentration (Pitoisuus).
  - Concentration (Pitoisuus): inputs: Päästö, LO, LA.
    - LO and LA: given within the ovariable emission (Päästö).
    - Uses function GIS.Concentration.matrix(Päästö, LO, LA, ...)
    - Emission (Päästö): inputs Traffic volume (Tieliikennepäästöt) (at this point, only unit conversions).
      - Traffic volumes (Tieliikennepäästöt): uses the method Tieliikennepäästöt and ovariable Activity (suorite).
        Activity (suorite): is given on the data table of the page Pienhiukkaspäästöt Raumalla.

Data

Health impacts of fine particles in Rauma(-)
Obs	City planning option	Vehicle	Mode	LA	LO	Unit	Traffic volume	Length	Description
1	Option 1	Truck	Road transport with 70 % load	61.131634	21.491318	kpl-km /d	2662	9.5	Emission location is asssumed near the centre of Rauma
2	Option 1	Electric train		61.131634	21.491318	ton-km /d	6953	9.5	Emission location is asssumed near the centre of Rauma
3	Option 1	Diesel train		61.131634	21.491318	ton-km /d	6953	9.5	Emission location is asssumed near the centre of Rauma
4	Option 2	Truck	Road transport with 70 % load	61.131634	21.491318	kpl-km /d	1246	9.5	Emission location is asssumed near the centre of Rauma
5	Option 2	Truck	Road transport with 70 % load	61.106698	21.492348	kpl-km /d	1653	10.1	Emission location is assumed at highway 8.
6	Option 2	Electric train		61.097802	21.492691	ton-km /d	2775	14.9	Emission location is assumed at the southern railroad.
7	Option 2	Diesel train		61.097802	21.492691	ton-km /d	2775	14.9	Emission location is assumed at the southern railroad.
8	Option 2	Electric train		61.131634	21.491318	ton-km /d	3254	9.5	Emission location is asssumed near the centre of Rauma
9	Option 2	Diesel train		61.131634	21.491318	ton-km /d	3254	9.5	Emission location is asssumed near the centre of Rauma
10	Option 3	Truck	Road transport with 70 % load	61.131634	21.491318	kpl-km /d	2662	9.5	Emission location is asssumed near the centre of Rauma
11	Option 3	Electric train		61.131634	21.491318	ton-km /d	6953	9.5	Emission location is asssumed near the centre of Rauma
12	Option 3	Diesel train		61.131634	21.491318	ton-km /d	6953	9.5	Emission location is asssumed near the centre of Rauma
13	Option 1	Ship		61.127095	21.407547	ton /a	5.7		Ship emissions, location is assumed at Ruuhiluoto
14	Option 2	Ship		61.121688	21.444626	ton /a	5.7		Ship emissions, location is assumed at Järviluoto
15	Option 3	Ship		61.127095	21.407547	ton /a	5.7		Ship emissions, location is assumed at Ruuhiluoto

Table contains the traffic that changes between the three city planning options in the port, city centre, or highway. The traffic volumes are assumed to correspond the port expansion plan Ve2 (Ruuhiluoto) in options 1 and 3; and port expansion plan Ve3 (Järviluoto) in option 2.
The lengths of routes and assumed emission points (that are chosen to reflect the typical emission distance from the population) are shown on the Google map below. Single points are used because the dispersion model does not take line sources as inputs.
Traffic volumes are taken from the Environmental Impact assessment of the Rauma Port Expansion^[1], from which the amounts are converted to daily basis.
- Half of the trains are assumed to be electric, half diesel.

**Emissions to air in different port expansion options ^[1]**
Ship emissions on sea and at port (t/a)
Port expansion option	CO	NOx	PM10	PM2.5	SO2	CO2
Current	14.8	200.7	3.1	2.5	28.9	10193.0
Ve0	17.4	235.6	3.7	2.9	33.9	11964.3
Ve1 (Saukot)	22.6	305.9	4.8	3.8	44.0	15538.1
Ve2 (Ruuhiluoto)	33.9	458.9	7.2	5.7	66.0	23307.2
Ve3 (Järviluoto)	33.9	458.9	7.2	5.7	66.0	23307.2
Ve4 (Hanskloppi)	31.6	428.3	6.7	5.4	61.6	21753.4
Heavy duty vehicles (t/a)
Port expansion option	CO	HC	NOx	PM	SO2	CO2
Current	2.9	1.7	34.3	0.6	0.0	4868.7
Ve0	3.3	1.9	40.1	0.7	0.1	5680.1
Ve1 (Saukot)	4.9	2.8	58.3	1.0	0.1	8260.3
Ve2 (Ruuhiluoto)	7.1	4.1	85.6	1.4	0.1	12135.2
Ve3 (Järviluoto)	7.8	4.5	93.2	1.6	0.1	13215.6
Ve4 (Hanskloppi)	7.4	4.3	89.2	1.5	0.1	12645.8

References

↑ ^1.0 ^1.1 Rauman sataman laajennuksen YVA-selostus

Related files

[raumayva-1] 1.0 ^1.1 Rauman sataman laajennuksen YVA-selostus

[1]