GSODR use case: Specified years/stations vignette

required <- c("raster", "rgdal", "rgeos")

if (!all(unlist(lapply(required, function(pkg) requireNamespace(pkg, quietly = TRUE)))))
    knitr::opts_chunk$set(eval = FALSE, collapse = TRUE, comment = "#>", fig.align = "center")

Gather Data for IRRI Central Luzon Survey Loop

The IRRI (International Rice Research Institute) survey loop in Central Luzon is a study that aims to monitor the changes in rice farming in the major rice producing area of the Philippines - the Central Luzon region, which is called as the "rice bowl of the Philippines". Data have been collected in this project since the 1960s. See, for the panel data.

This vignette details how to find and retrieve weather data for the area that this survey covers for the time period of 1960-2016. Methods that are detailed include:

Retrieve PHL Provincial Data and Select Loop Provinces

As a first step, we'll use the raster package to retrieve data from GADM that will provide the provincial spatial data for the survey area. We will then use this to find the centroid of the area of interest, which will be used to find the nearest stations. Using raster::getData() fetch level 0 (national) and 1 (provincial) data for the Philippines.

RP0 <- raster::getData(country = "Philippines", level = 0)
RP1 <- raster::getData(country = "Philippines", level = 1)

Now we will select the provinces involved in the survey and make a new object called Central_Luzon from the provincial level data, RP1.

Central_Luzon <- RP1[RP1@data$NAME_1 == "Pampanga" | 
                     RP1@data$NAME_1 == "Tarlac" |
                     RP1@data$NAME_1 == "Pangasinan" |
                     RP1@data$NAME_1 == "La Union" |
                     RP1@data$NAME_1 == "Nueva Ecija" |
                     RP1@data$NAME_1 == "Bulacan", ]

Next, create a map inset showing where the Central Luzon Loop survey takes place.

First we'll use gSimplify() from rgeos to simplify the map of the Philippines to make the map generation in the next few steps quicker.

RP0 <- rgeos::gSimplify(RP0, tol = 0.05)

CL_names <- data.frame(coordinates(Central_Luzon)) # get center coordinates of provinces in Central Luzon
CL_names$label <- Central_Luzon@data$NAME_1

# Main Map
p1 <- ggplot() + 
  geom_polygon(data = Central_Luzon, aes(x = long, y = lat, group = group),
               colour = "grey10", fill = "#fff7bc") +
  geom_text(data = CL_names, aes(x = X1, y = X2, label = label), 
            size = 2, colour = "grey20") +
  theme(axis.text.y = element_text(angle = 90, hjust = 0.5)) +
  ggtitle("Central Luzon Provinces Surveyed") +
  theme_bw() + 
  xlab("Longitude") + 
  ylab("Latitude") +

# Inset
p2 <- ggplot() + 
  geom_polygon(data = RP0, aes(long, lat, group = group),
               colour = "grey10",
               fill = "#fff7bc") +
  coord_equal() +
  theme_bw() + 
  labs(x = NULL, y = NULL) +
  geom_rect(aes(xmin = extent(Central_Luzon)[1],
                xmax = extent(Central_Luzon)[2],
                ymin = extent(Central_Luzon)[3],
                ymax = extent(Central_Luzon)[4]),
            alpha = 0,
            colour = "red",
            size = 0.7,
            linetype = 1) +
  theme(axis.text.x = element_blank(),
        axis.text.y = element_blank(),
        axis.ticks = element_blank(),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        plot.margin = unit(c(0, 0, 0 ,0), "mm"))

v1 <- viewport(width = 1, height = 1, x = 0.5, y = 0.5) # plot area for the main map
v2 <- viewport(width = 0.28, height = 0.28, x = 0.74, y = 0.79) # plot area for the inset map
print(p1, vp = v1) 
print(p2, vp = v2)

Map of provinces surveyed

Dissolve Polygons and Find Centroid of Loop Survey Area

Now that we have the provincial data that we want, we will dissolve the polygons that represent the individual provinces in Central Luzon and find the centroid of all of them, which we will use as the central point for querying stations from the GSOD data set.

Central_Luzon <- rgeos::gUnaryUnion(Central_Luzon)
centroid <- rgeos::gCentroid(Central_Luzon)

ggplot() + 
  geom_polygon(data = Central_Luzon, aes(x = long, y = lat, group = group),
               colour = "grey10", fill = "#fff7bc") +
  geom_point(aes(x = centroid@coords[1], y = centroid@coords[2])) +
  theme(axis.text.y = element_text(angle = 90, hjust = 0.5)) +
  ggtitle("Centre of Survey Area") +
  theme_bw() + 
  xlab("Longitude") + 
  ylab("Latitude") +

Map of disolved polygons with centroid point

Next, make a list of stations that are within this area. First we need to fetch the station metadata, "isd-history.csv" from the FTP server and then check which stations fall within a 100km radius of the centre of the provinces we are interested in.

# Fetch station list from NCEI
station_meta <- read_csv(
  col_types = "ccccccddddd",
  col_names = c("USAF", "WBAN", "STN_NAME", "CTRY", "STATE", "CALL", "LAT",
                "LON", "ELEV_M", "BEGIN", "END"), skip = 1)
station_meta$STNID <- as.character(paste(station_meta$USAF,
                                         sep = "-"))

loop_stations <- nearest_stations(LAT = centroid@coords[2],
                                  LON = centroid@coords[1], 
                                  distance = 100)

loop_stations <- station_meta[station_meta$STNID %in% loop_stations, ]

loop_stations <- loop_stations[loop_stations$BEGIN <= 19591231 &
                               loop_stations$END >= 20151231, ]

print(loop_stations[, c(1:2, 3, 7:12)])
# A tibble: 4 x 9
  USAF   WBAN  STN_NAME           LAT   LON  ELEV_M     BEGIN       END STNID       
  <chr>  <chr> <chr>            <dbl> <dbl>   <dbl>     <dbl>     <dbl> <chr>       
1 983240 99999 IBA/LUZON ISLAND  15.3  120.    5.00 19490104. 20180327. 983240-99999
2 983250 99999 DAGUPAN           16.1  120.    2.00 19450119. 20180327. 983250-99999
3 983270 99999 CLARK INTL        15.2  121.  148.   19450214. 20180327. 983270-99999
4 983280 99999 BAGUIO            16.4  121. 1296.   19490101. 20180327. 983280-99999

These are all of the stations that are available within 100km of the centroid of this area and the years for which data are available.

Using get_GSOD() to Fetch the Requested Station Files

This example shows how you could construct a query using the get_GSOD() function. Be aware that it may result in incomplete data and error from the server if it stops responding. We've done our best to make GSODR handle these errors, but if it does this, see the following option for using the reformat_GSOD() function.

PHL <- get_GSOD(station = 
                  eval(parse(text = loop_stations[, 12])), years = 1960:2015)

Another Option, Using reformat_GSOD()

GSODR provides a function for dealing with local files that have been transferred from the server already as well, reformat_GSOD(). If the previous example with get_GSOD() does not work, this is a good alternative that takes a bit more intervention but gives the same results.

Using your FTP client (e.g., FileZilla) log into the NCEI FTP server, and navigate to /pub/data/gsod/. Manually downloading the files for each station listed above from 1960 to 2015 is possible, but tedious. An easier solution is to simply download the annual files found in each yearly directory, "gsod-YYYY.tar" and untar them locally and then use R to list the available files and select only the files for the stations of interest. Lastly, write the data to disk as a CSV file for saving and later use.

years <- 1960:2015

loop_stations <- eval(parse(text = loop_stations[, 12]))

# create file list
loop_stations <-
  paste0, c(expand.grid(loop_stations, "-", years, ".op.gz"))

local_files <- list.files(path = "./GSOD", full.names = TRUE, recursive = TRUE)
local_files <- local_files[basename(local_files) %in% loop_stations]

loop_data <- reformat_GSOD(file_list = local_files)

readr::write_csv(loop_data, path = "Loop_Survey_Weather_1960-2015", path = "./")



Elevation Values

90m hole-filled SRTM digital elevation (Jarvis et al. 2008) was used t o identify and correct/remove elevation errors in data for station locations between -60˚ and 60˚ latitude. This applies to cases here where elevation was missing in the reported values as well. In case the station reported an elevation and the DEM does not, the station reported is taken. For stations beyond -60˚ and 60˚ latitude, the values are station reported values in every instance. See for more detail on the correction methods.

WMO Resolution 40. NOAA Policy

Users of these data should take into account the following (from the NCEI website):

"The following data and products may have conditions placed on their international commercial use. They can be used within the U.S. or for non-commercial international activities without restriction. The non-U.S. data cannot be redistributed for commercial purposes. Re-distribution of these data by others must provide this same notification." WMO Resolution 40. NOAA Policy


Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E. (2008) Hole-filled SRTM for the globe Version 4, available from the CGIAR-CSI SRTM 90m Database (

Try the GSODR package in your browser

Any scripts or data that you put into this service are public.

GSODR documentation built on June 16, 2018, 1:38 p.m.