rainfall
Otero, Kate Andrea
2023-07-13
importing data
rainfall <- read.csv("RAINFALL DATA.csv")plot GPM_IMERG data
#convert into time series
#Maasin
GPM.ts.msn <- ts(rainfall$GPM.IMERG.Maasin.Rainfall,
start = c(2010,1),
end = c(2020,365),
frequency = 365)
autoplot(GPM.ts.msn,
ylab = "Rainfall Amount",
xlab = "Time") +
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"))
#Tacloban
GPM.ts.tac <- ts(rainfall$GPM.IMERG.Tacloban.Rainfall,
start = c(2010,1),
end = c(2020,365),
frequency = 365)
autoplot(GPM.ts.tac,
ylab = "Rainfall Amount",
xlab = "Time") +
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"))
#VSU
#Tacloban
GPM.ts.vsu <- ts(rainfall$GPM.IMERG.VSU.Rainfall,
start = c(2010,1),
end = c(2020,365),
frequency = 365)
autoplot(GPM.ts.vsu,
ylab = "Rainfall Amount",
xlab = "Time") +
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"))
#ploting GPM vs in-situ
#converting in-situ data into time series
#Maasin
Msn_nc2 <- ts(rainfall$Maasin.Rainfall,
start = c(2010,1),
end = c(2020,365),
frequency = 365)
autoplot(GPM.ts.msn)+
autolayer(Msn_nc2)+
ylab("Rainfall Amount")+
xlab("Time")+
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
legend.position="none",
text = element_text(size = 12, family = "TT Times New Roman"))## Warning in grid.Call(C_stringMetric, as.graphicsAnnot(x$label)): font family not
## found in Windows font database
## Warning in grid.Call(C_stringMetric, as.graphicsAnnot(x$label)): font family not
## found in Windows font database
## Warning in grid.Call(C_stringMetric, as.graphicsAnnot(x$label)): font family not
## found in Windows font database## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
#Tacloban
Tac_nc2 <- ts(rainfall$Tacloban.Rainfall,
start = c(2010,1),
end = c(2020,365),
frequency = 365)
autoplot(GPM.ts.tac)+
autolayer(Tac_nc2)+
ylab("Rainfall Amount")+
xlab("Time")+
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
legend.position="none",
text = element_text(size = 12, family = "TT Times New Roman"))## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
#VSU
VSU_nc2 <- ts(rainfall$VSU.Rainfall,
start = c(2010,1),
end = c(2020,365),
frequency = 365)
autoplot(GPM.ts.vsu)+
autolayer(VSU_nc2)+
ylab("Rainfall Amount")+
xlab("Time")+
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
legend.position="none",
text = element_text(size = 12, family = "TT Times New Roman"))## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database## Warning in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : font
## family not found in Windows font database
univariate linear fitting
#partitioning the data set
ind = sample.split(Y = rainfall$GPM.IMERG.Maasin.Rainfall, SplitRatio = 0.7)
#subsetting into Train data
train <- rainfall[ind,]
#subsetting into Test data
test <- rainfall[!ind,]
#MAASIN
#correlation of satellite vs ground data
cor.test(rainfall$Maasin.Rainfall,rainfall$GPM.IMERG.Maasin.Rainfall)##
## Pearson's product-moment correlation
##
## data: rainfall$Maasin.Rainfall and rainfall$GPM.IMERG.Maasin.Rainfall
## t = 19.274, df = 4016, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.2624222 0.3190343
## sample estimates:
## cor
## 0.2909829#fitting regression model
msn_lm.u <- lm(GPM.IMERG.Maasin.Rainfall ~ Maasin.Rainfall, data = train)
summary(msn_lm.u)##
## Call:
## lm(formula = GPM.IMERG.Maasin.Rainfall ~ Maasin.Rainfall, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -64.000 -6.259 -5.692 -0.079 230.856
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 6.25909 0.38257 16.36 <2e-16 ***
## Maasin.Rainfall 0.29001 0.01783 16.27 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.8 on 2810 degrees of freedom
## Multiple R-squared: 0.08609, Adjusted R-squared: 0.08577
## F-statistic: 264.7 on 1 and 2810 DF, p-value: < 2.2e-16ggplot(train, aes(x=Maasin.Rainfall, y=GPM.IMERG.Maasin.Rainfall))+
geom_point(size=0.5)+ geom_smooth(method=lm)+
labs(x="Rainfall Data", y="GPM IMERG")+
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"))## `geom_smooth()` using formula = 'y ~ x'
#TACLOBAN
#correlation of satellite vs ground data
cor.test(rainfall$Tacloban.Rainfall,rainfall$GPM.IMERG.Tacloban.Rainfall)##
## Pearson's product-moment correlation
##
## data: rainfall$Tacloban.Rainfall and rainfall$GPM.IMERG.Tacloban.Rainfall
## t = 25.913, df = 3743, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.3625161 0.4168391
## sample estimates:
## cor
## 0.3900169#fitting regression model
tac_lm.u <- lm(GPM.IMERG.Tacloban.Rainfall ~ Tacloban.Rainfall, data = train)
summary(tac_lm.u)##
## Call:
## lm(formula = GPM.IMERG.Tacloban.Rainfall ~ Tacloban.Rainfall,
## data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -88.35 -6.19 -5.65 -0.14 316.54
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 6.07579 0.43664 13.91 <2e-16 ***
## Tacloban.Rainfall 0.41957 0.01854 22.63 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 20.89 on 2623 degrees of freedom
## (187 observations deleted due to missingness)
## Multiple R-squared: 0.1633, Adjusted R-squared: 0.163
## F-statistic: 512 on 1 and 2623 DF, p-value: < 2.2e-16ggplot(train, aes(x=Tacloban.Rainfall, y=GPM.IMERG.Tacloban.Rainfall))+
geom_point(size=0.5)+ geom_smooth(method=lm)+
labs(x="Rainfall Data", y="GPM IMERG")+
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"))## `geom_smooth()` using formula = 'y ~ x'## Warning: Removed 187 rows containing non-finite values (`stat_smooth()`).## Warning: Removed 187 rows containing missing values (`geom_point()`).
#VSU
#correlation of satellite vs ground data
cor.test(rainfall$VSU.Rainfall,rainfall$GPM.IMERG.VSU.Rainfall)##
## Pearson's product-moment correlation
##
## data: rainfall$VSU.Rainfall and rainfall$GPM.IMERG.VSU.Rainfall
## t = 19.47, df = 4015, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.2652085 0.3217285
## sample estimates:
## cor
## 0.2937252#fitting regression model
vsu_lm.u <- lm(GPM.IMERG.VSU.Rainfall ~ VSU.Rainfall, data = train)
summary(vsu_lm.u)##
## Call:
## lm(formula = GPM.IMERG.VSU.Rainfall ~ VSU.Rainfall, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -60.066 -6.342 -5.880 -0.105 251.814
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 6.34242 0.39598 16.02 <2e-16 ***
## VSU.Rainfall 0.29135 0.01706 17.08 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.24 on 2810 degrees of freedom
## Multiple R-squared: 0.09402, Adjusted R-squared: 0.09369
## F-statistic: 291.6 on 1 and 2810 DF, p-value: < 2.2e-16ggplot(train, aes(x=VSU.Rainfall, y=GPM.IMERG.VSU.Rainfall))+
geom_point(size=0.5)+ geom_smooth(method=lm)+
labs(x="Rainfall Data", y="GPM IMERG")+
theme(panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"))## `geom_smooth()` using formula = 'y ~ x'
#prediction
msn_predict <- 6.5191 + 0.2875*test$Maasin.Rainfall
tac_predict <- 6.1398 + 0.3936*test$Tacloban.Rainfall
vsu_predict <- 6.2288 + 0.2888*test$VSU.Rainfall
df <- data.frame(msn_predict,test$GPM.IMERG.Maasin.Rainfall,tac_predict,test$GPM.IMERG.Tacloban.Rainfall,vsu_predict,test$GPM.IMERG.VSU.Rainfall)
write.csv(df, "C:/Users/Admin/OneDrive/Desktop/rainfall\\univariate_linear_fitting.csv")#multivariate linear fitting
#fitting regression model
#cleaning data values in temperature
rainfall[rainfall == "-999"] <- NA
rainfall <- na.omit(rainfall)
train[train == "-999"] <- NA
train <- na.omit(train)
#MAASIN
msn_lm.m <- lm(GPM.IMERG.Maasin.Rainfall ~ Maasin.Rainfall + MAASIN.TMEAN + MAASIN.RH, data = train)
summary(msn_lm.m)##
## Call:
## lm(formula = GPM.IMERG.Maasin.Rainfall ~ Maasin.Rainfall + MAASIN.TMEAN +
## MAASIN.RH, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -50.245 -7.358 -4.010 0.538 170.241
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 50.41689 14.98124 3.365 0.000777 ***
## Maasin.Rainfall 0.17371 0.02075 8.373 < 2e-16 ***
## MAASIN.TMEAN -2.07904 0.33584 -6.191 7.03e-10 ***
## MAASIN.RH 0.15439 0.09242 1.670 0.094960 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.68 on 2401 degrees of freedom
## Multiple R-squared: 0.0962, Adjusted R-squared: 0.09507
## F-statistic: 85.19 on 3 and 2401 DF, p-value: < 2.2e-16#TACLOBAN
tac_lm.m <- lm(GPM.IMERG.Tacloban.Rainfall ~ Tacloban.Rainfall + TAC.TMEAN + TAC.RH, data = train)
summary(tac_lm.m)##
## Call:
## lm(formula = GPM.IMERG.Tacloban.Rainfall ~ Tacloban.Rainfall +
## TAC.TMEAN + TAC.RH, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -77.10 -6.88 -4.55 -0.57 318.13
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -29.91688 18.80321 -1.591 0.111730
## Tacloban.Rainfall 0.34885 0.02159 16.161 < 2e-16 ***
## TAC.TMEAN 0.24849 0.46473 0.535 0.592913
## TAC.RH 0.34983 0.09858 3.549 0.000395 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 21 on 2401 degrees of freedom
## Multiple R-squared: 0.1432, Adjusted R-squared: 0.1421
## F-statistic: 133.7 on 3 and 2401 DF, p-value: < 2.2e-16#VSU
vsu_lm.m <- lm(GPM.IMERG.VSU.Rainfall ~ VSU.Rainfall + VSU.TMEAN + VSU.RH, data = train)
summary(vsu_lm.m)##
## Call:
## lm(formula = GPM.IMERG.VSU.Rainfall ~ VSU.Rainfall + VSU.TMEAN +
## VSU.RH, data = train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -50.326 -7.260 -4.220 -0.011 251.139
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 31.47263 16.30286 1.930 0.053663 .
## VSU.Rainfall 0.19155 0.02007 9.543 < 2e-16 ***
## VSU.TMEAN -2.03934 0.41038 -4.969 7.19e-07 ***
## VSU.RH 0.38489 0.09901 3.887 0.000104 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.01 on 2401 degrees of freedom
## Multiple R-squared: 0.09747, Adjusted R-squared: 0.09634
## F-statistic: 86.43 on 3 and 2401 DF, p-value: < 2.2e-16#multivariate prediction
msn.m_predict <- 45.4932 + (0.1419*test$Maasin.Rainfall) - (1.9930*test$MAASIN.TMEAN) + (0.1867*test$MAASIN.RH)
tac.m_predict <- 0.8472 + (0.3594*test$Tacloban.Rainfall) - (0.2176*test$TAC.RH)
vsu.m_predict <- 39.2318 + (0.1769*test$VSU.Rainfall) -(2.2382*test$VSU.TMEAN) + 0.3563*test$VSU.RH
df <- data.frame(msn.m_predict,test$GPM.IMERG.Maasin.Rainfall,tac.m_predict,test$GPM.IMERG.Tacloban.Rainfall,vsu.m_predict,test$GPM.IMERG.VSU.Rainfall)
write.csv(df, "C:/Users/Admin/OneDrive/Desktop/rainfall\\multivariate_linear_fitting.csv")#plotting multivariate regrssion results
avPlots(msn_lm.m, layout = c(1,3), grid=F, ylab="GPM IMERG", main = "", )
avPlots(tac_lm.m, layout = c(1,3), grid=F, ylab="GPM IMERG", main = "")
avPlots(vsu_lm.m, layout = c(1,3), grid=F, ylab="GPM IMERG", main = "")