New York Times journalist Justin Gillis has published a series of articles on the impact of Global Warming. The most recent one appeared on March 31. It chronicles the view of many scientists that the Antacrctica ice sheet is melting and the profound effect this will have on sea levels. What is striking about this analysis is the claim that the melting process will take place over the course of decades, not centuries.

http://www.nytimes.com/2016/03/31/science/global-warming-antarctica-ice-sheet-sea-level-rise.html?_r=0

If these predictions are true, what are the consequences to global popultion centers in coastal and low-lying areas globally? This is what the following exercise attempts to illustrate using geo and semantic web-scraping techniques provided by the NY Times API.

rinterface

rinterface

0.1 Search on articles by Justin Gillis

semantic_key <- "&api-key=621746db4c8f966bdf1f67dca6aa3b90:13:65002889"
url <- "http://api.nytimes.com/svc/semantic/v2/concept/name/nytd_per/Gillis,%20Justin.json?fields=all"
json_file <- paste0(url, semantic_key)
json_data_raw <- fromJSON(json_file)

#review meta-data 
df_nyt_meta <- data.frame(
  status = c(unlist(json_data_raw[1])),
  copyright = c(unlist(json_data_raw[2])),
  num_results = c(unlist(json_data_raw[3]))
)

#display meta-data
kable(df_nyt_meta, align = 'l')
status copyright num_results
status OK Copyright (c) 2015 The New York Times Company. All Rights Reserved. 1 
kable(data.frame(Metadata=names(json_data_raw$results)), align = 'l')

0.2 Metadata

concept_id
concept_name
is_times_tag
concept_status
vernacular
concept_type
concept_created
concept_updated
teragram
variants
pages
search_api_query article_list

0.3 Extracts of content from articles

#get interesting info from sublists
df_article_bodies <- data.frame(Excerpts = json_data_raw$results$article_list$results[[1]][[1]])
df_article_nytd_des <- data.frame(article_desc = unlist(json_data_raw$results$article_list$results[[1]][[3]]$nytd_des))
df_article_nytd_geo <- data.frame(article_geo = unlist(json_data_raw$results$article_list$results[[1]][[3]]$nytd_geo))
df_results<-data.frame(json_data_raw[['results']])
df_geocodes <- data.frame(json_data_raw[[4]][[11]])

#display article bodies by Justin Gillis on global warming.
kable(df_article_bodies, align='l')

0.4 Excerpts

A host of approaches to limiting global warming emerge as world leaders gather in Paris to launch two weeks of climate talks.
A powerful El Niño and greenhouse-driven warming combine to push 2015 into the record books.
A new study of Antarctic ice and warming finds humanity will have to write off today’s coastal cities centuries from now if it chooses to keep burning fossil fuels.
The latest global warming report from the U.N. climate panel helpfully clarifies risks but largely misses a key factor in driving an energy transformation — research. Adam Bryant recently became editor of The Times’s expanded team covering the environment. We asked him how he is approaching the position.
The new U.N. climate panel report on impacts of global warming sees substantial shifts already under way and much more to come.
A closer look at the science behind the fighting over whether global warming has shaped California’s drought.
A new modeling study finds that by mid-century many places on Earth, starting in the tropics, will have a fundamentally warmer climate.
A fresh look at studies finding a less potent warming from the continuing buildup of greenhouse gases.
A famous curve of carbon dioxide levels is close to passing a milestone.

0.5 World cities by elevation & population

loc <- geo_search(feature_class = 'P')
df_location <- data.frame(loc)
df_location[35,12] <- 8400000 # population new york city
df_location[35,11] <- 33 # elevation new york city
df_location[33,11] <- 6 # elevation Shanghai
df_location[46,11] <- 25 # elevation Buenos Aires
df_location[47,11] <- 14 # elevation Mumbai
df_location[31,11] <- 2250 # elevation Mexico City
df_location[53,11] <- 40 # elevation Istanbul
df_location[22,11] <- 7786 # elevation Moscow

df_location <- df_location[order(-df_location$data.population),]
df_location['size'] <- df_location$data.population/5000000
kable(head(df_location[,c(5,11,12)], n=10), align='l')
data.concept_name data.elevation data.population
33 Shanghai (China) 6 14608512
46 Buenos Aires (Argentina) 25 13076300
47 Mumbai (India) 14 12691836
31 Mexico City (Mexico) 2250 11285654
53 Istanbul (Turkey) 40 11174257
22 Moscow (Russia) 7786 10381222
42 Seoul (South Korea) NA 10349312
88 Jakarta (Indonesia) NA 8540121
35 Times Square and 42nd Street (NYC) 33 8400000
18 Tokyo (Japan) NA 8336599

0.6 World cities vulnerability to flooding

#change the location data into a SpatialPointsDataFrame
coords <- cbind(Longitude = as.numeric(as.character(df_location$data.longitude)), 
                Latitude = as.numeric(as.character(df_location$data.latitude)))
df_location.pts <- SpatialPointsDataFrame(coords, df_location, proj4string = CRS("+init=epsg:4326"))
#plot(df_location.pts, pch = ".", col = "darkred")

mp <- NULL
mapWorld <- borders("world", colour="gray50", fill="gray60") # create a layer of borders
mp <- ggplot() +   mapWorld

#Now Layer the cities on top
mp <- mp + geom_point(aes(x=df_location$data.longitude, y=df_location$data.latitude, color=df_location$data.elevation), size=df_location$size)
mp <- mp +  scale_colour_gradient(limits=c(0, 200), low="red", na.value = "antiquewhite", guide = "colourbar")  
mp
## Warning: Removed 1 rows containing missing values (geom_point).

http://www.agcs.allianz.com/assets/PDFs/GRD/GRD%20individual%20articles/022014/SpecialTopicEarth_the-heat-is-on.pdf rinterface

0.7 US cities vulnerability to flooding

df_location <- subset(df_location, data.country_code == 'US')
df_location <- df_location[df_location$data.name != "Honolulu", ]
df_location['size'] <- df_location$data.population/1000000
kable(head(df_location[,c(5,11,12)], n=10), align='l')
data.concept_name data.elevation data.population
35 Times Square and 42nd Street (NYC) 33 8400000
1 Los Angeles (Calif) 89 3694820
3 Chicago (Ill) 179 2841952
13 Houston (Tex) 12 2027712
14 Philadelphia (Pa) 12 1517550
52 Phoenix (Ariz) 331 1321045
32 San Diego (Calif) 20 1223400
39 Dallas (Tex) 128 1211704
10 Detroit (Mich) 183 951270
5 San Francisco (Calif) 16 732072 
#change the location data into a SpatialPointsDataFrame
coords <- cbind(Longitude = as.numeric(as.character(df_location$data.longitude)), 
                Latitude = as.numeric(as.character(df_location$data.latitude)))
df_location.pts <- SpatialPointsDataFrame(coords, df_location, proj4string = CRS("+init=epsg:4326"))
#plot(df_location.pts, pch = ".", col = "darkred")

mp <- NULL
mapUSA <- borders('state', colour="gray50", fill="gray60")
mp <- ggplot() +   mapUSA

#Now Layer the cities on top
mp <- mp + geom_point(aes(x=df_location$data.longitude, y=df_location$data.latitude, color=df_location$data.elevation), size=df_location$size)
mp <- mp + scale_colour_gradient(limits=c(0, 200), low="red", na.value = "antiquewhite", guide = "colourbar")
mp
## Warning: Removed 1 rows containing missing values (geom_point).

0.8 NY cities vulnerability to flooding

df_location <- subset(df_location, data.admin_code1 == 'NY')
df_location['size'] <- df_location$data.population/1000000
kable(df_location[,c(5,11,12)], align='l')
data.concept_name data.elevation data.population
35 Times Square and 42nd Street (NYC) 33 8400000
71 Buffalo (NY) 183 279557
38 Yonkers (NY) 25 197818
15 Albany (NY) 45 95658
50 White Plains (NY) 65 57260
89 Bedford-Stuyvesant (NYC) 15 NA 
coords <- cbind(Longitude = as.numeric(as.character(df_location$data.longitude)), 
                Latitude = as.numeric(as.character(df_location$data.latitude)))
df_location.pts <- SpatialPointsDataFrame(coords, df_location, proj4string = CRS("+init=epsg:4326"))
#plot(df_location.pts, pch = ".", col = "darkred")

mp <- NULL
mapNY <- borders("county","new york", colour="gray50", fill="gray60") 
mp <- ggplot() +   mapNY

#Now Layer the cities on top
mp <- mp + geom_point(aes(x=df_location$data.longitude, y=df_location$data.latitude, color=df_location$data.elevation), size=df_location$size)
mp <- mp + scale_colour_gradient(limits=c(0, 200), low="red", na.value = "antiquewhite", guide = "colourbar")
mp
## Warning: Removed 1 rows containing missing values (geom_point).

0.9 Implications

What do the implications of severe flooding look like for a major city like New York? The following study was conducted by Hunter College using data collected by FEMA.

Mr. Gillis’s article reports that the prevailing view of scientists studying the the melting of the Antarctic ice shelf is that sea-levels will rise an average of 3 feet by the end of the century, accelerating mid-century. Note that many of the most highly populated cities are along coastal areas and are within flood surge range of these projections.

http://carsi.hunter.cuny.edu/project/hurricane-sandy-3d-flood-map/ rinterface

rinterface

rinterface