Extract Storms

In StormR: Analyzing the Behaviour of Wind Generated by Tropical Storms and Cyclones

``` {r setup, include = FALSE} library(StormR) library(terra) library(rworldxtra) data("countriesHigh")

The `defStormsList()` function allows to extract tropical cyclone track data for a
given tropical cyclone or set of tropical cyclones nearby a given location of
interest (`loi`). The `loi` can be defined using a country name, a specific
point (defined by its longitude and latitude coordinates), or any user imported
or defined spatial polygon shapefiles. By default only observations located
within 300 km around the `loi` are extracted but this can be changed using the
`max_dist` argument. Users can also extract tropical cyclones using the `name`
of the storm or the `season` during which it occurred. If both the `name` and
the `season` arguments are not filled then the `defStormsList()` function extracts all
tropical cyclones since the first cyclonic season in the database. Once the data
are extracted, the `plotStorms()` function can be used to visualize the
trajectories and points of observation of extracted tropical cyclones on a map.

In the following example we use the `test_dataset` provided with the package to
illustrate how cyclone track data can be extracted and visualised using country
and cyclone names, specific point locations, and polygon shapefiles, as
described below.

### Getting and ploting tropical cyclone track data

#### Using country names

We extract data on the tropical cyclone Pam (2015) nearby Vanuatu as follows:  

``` {r chunk1}
sds <- defStormsDataset(verbose = 0)
st <- defStormsList(sds = sds, loi = "Vanuatu", names = "PAM", verbose = 0)

The defStormsList() function returns a stormsList object in which the first slot @data contains a list of Storm objects. With the above specification the stormsList contains only one Storm object corresponding to cyclone PAM and the track data can be obtained using the getObs() function as follows:

``` {r chunk2} head(getObs(st, name = "PAM"))

The number of observation and the indices of the observations can be obtained
using the `getNbObs()` and `getInObs()` as follows:

``` {r chunk3}
getNbObs(st, name = "PAM")
getInObs(st, name = "PAM")

The data can be visualised on a map as follows:

``` {r chunk4} plotStorms(st, labels = TRUE)

#### Using a specified point location

We can extract all tropical cyclones near Nouméa (longitude = 166.45, latitude =
-22.27) between 2015 and 2021 as follows:

``` {r chunk5}
pt <- c(166.45, -22.27)
st <- defStormsList(sds = sds, loi = pt, seasons = c(2015, 2021), verbose = 0)

The number, the names, and the season of occurrence of the storms in the returned stormsList object can be obtained using the getNbStorms(), getNames(), and getSeasons() functions as follows:

``` {r chunk6} getNbStorms(st) getNames(st) getSeasons(st)

We can plot track data for the topical cyclone Niran only using the `names`
argument of the `plotStorms()` function as follows:

``` {r chunk7}
plotStorms(st, names = "NIRAN", labels = TRUE, legends = "bottomleft")

The track data for Niran can also be extracted and stored in a new object using the getStorm() function as follows:

``` {r chunk8} NIRAN <- getStorm(st, name = "NIRAN") getNames(NIRAN)

#### Using a user defined spatial polygon shapefile 

We can extract all tropical cyclones that occurred between 2015 and 2021 near
the New Caledonia exclusive economic zone using the `eezNC` shapefile provided
with the `StormR` package as follows:

``` {r chunk9}
sp <- eezNC
st <- defStormsList(sds = sds, loi = eezNC, season = c(2015, 2021), verbose = 0)

Information about the spatial extent of the track data exaction can be obtained using the getLOI(), getBuffer(), and getBufferSize() functions as follows:

``` {r chunk10} LOI <- getLOI(st) Buffer <- getBuffer(st) BufferSize <- getBufferSize(st) terra::plot(Buffer, lty = 3, main = paste(BufferSize, "km buffer arround New Caledonian EEZ", sep = " ")) terra::plot(LOI, add = TRUE) terra::plot(countriesHigh, add = TRUE)

### Using different wind scale


By default the Saffir-Simpson hurricane wind scale (SSHS) is used in `defStormsList()`
to assign level to storms.

The maximum level reached in the scale for each cyclone
can then be obtained using the `getScale()` function as follows:

``` {r chunk11}
getScale(st)

In this case, the SSHS scale is composed of 6 thresholds resulting in 6 levels spanning from level 0 to level 6.

We can only plot cyclones that reached level 5 and 6 using the category argument of the plotStorms() function as follows:

``` {r chunk12} plotStorms(st, category = c(5, 6), labels = FALSE, legends = "topright")

Finally, the user can choose his own scale and associated palette, by setting the `scale` and `scalePalette` inputs in `defStormsList()`. In the following example, we use the Tokyo's tropical cyclone intensity scale to analyse tropical storm PAM.

StormR provides default palette and category names:

```r
# Tokyo's tropical cyclone intensity scale
RSMCScale <- c(16.94, 24.44, 32.5, 43.33, 53.61)

sts_jpn <- defStormsList(sds = sds,
                         loi = "Vanuatu",
                         names = "PAM",
                         scale = RSMCScale,
                         verbose = 0)

plotStorms(sts_jpn)

But you can also easily customize them:

RSMCPalette <- c("#6ec1ea", "#4dffff", "#c0ffc0", "#ffd98c", "#ff738a", "#a188fc")
names(RSMCPalette) <- c("Tropical depression",
                        "Tropical storm",
                        "Severe tropical storm",
                        "Typhoon",
                        "Very strong typhoon",
                        "Violent typhoon")

sts_jpn <- defStormsList(sds = sds,
                         loi = "Vanuatu",
                         names = "PAM",
                         scale = RSMCScale,
                         scalePalette = RSMCPalette,
                         verbose = 0)

plotStorms(sts_jpn)

Dynamic plot

plotStorms allows the user to dynamically plot tracks within an interactive map using leaflet library by setting dynamicPlot to TRUE. Doing so, the user can explore the map the way he wants and click and each dotted colored observations to see there informations.

``` {r chunk13}

Example of dynamic plot, using the same parameters above

plotStorms(st, category = c(4, 5), labels = FALSE, legends = "topright", dynamicPlot=TRUE) ```

StormR documentation built on Sept. 12, 2024, 6:52 a.m.