solar navigation

November 9, 2014
By

(This article was first published on Dan Kelley Blog/R, and kindly contributed to R-bloggers)

Introduction

Solar altitude is a function of time, longitude and latitude, and so it can be
possible to infer location based on measuring altitude as a function of time.
This form of solar navigation can be based on sunrise and sunset times, at
least on non-equinox days.

I have explored this for a school-based project I call “SkyView” [1] involving
light sensors and Arduino microcontrollers, and I suspect that readers could
imagine other applications as well.

I will illustrate the ideas and the accuracy of the method based on the example
of sunrise and sunset on Remembrance Day in Halifax, Nova Scotia, a city where
respect for the fallen is very strong.

Methods

According to various websites [e.g. 2], sunrise on the Halifax Remembrance Day
sunrise is at 7:06AM (11:06 UTC), with sunset at 4:51PM (20:51 UTC).

Rather than devising an inverse formula to infer location from time and solar
altitude, the R function optim may be used to find the longitude and
latitude that minimize angle the sun makes to the horizon. That angle is given
by the altitude component of the list returned by oce::solarAngle().

Thus, a method for inferring the location of Halifax is as follows. The code
should be self-explanatory to anyone who can read R.

1
library(oce)
## Loading required package: methods
## Loading required package: mapproj
## Loading required package: maps
## Loading required package: proj4
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
misfit <- function(lonlat)
{
    riseAlt <- sunAngle(rise, longitude=lonlat[1], latitude=lonlat[2], useRefraction=TRUE)$altitude
    setAlt <- sunAngle(set, longitude=lonlat[1], latitude=lonlat[2], useRefraction=TRUE)$altitude
    0.5 * (abs(riseAlt) + abs(setAlt))
}
start <- c(-50, 50) # set this vaguely near the expected location
rise <- as.POSIXct("2014-11-11 11:06:00", tz="UTC")
set <- as.POSIXct("2014-11-11 20:51:00", tz="UTC")
bestfit <- optim(start, misfit)

# Plot coastline
data(coastlineWorldFine, package="ocedata")
plot(coastlineWorldFine, clon=-64, clat=45, span=500)
grid()

# Plot a series of points calculated by perturbing the 
# suggested times by about the rounding interval of 1 minute.
n <- 25
rises <- rise + rnorm(n, sd=30)
sets <- set + rnorm(n, sd=30)
set.seed(20141111) # this lets readers reproduce this exactly
for (i in 1:n) {
    rise <- rises[i]
    set <- sets[i]
    fit <- optim(start, misfit)
    points(fit$par[1], fit$par[2], pch=21, bg="pink", cex=1.4)
}
# Show the unperturbed fit
points(bestfit$par[1], bestfit$par[2], pch=21, cex=2, bg="red")
# Show the actual location of Halifax
points(-63.571, 44.649, pch=22, cex=2, bg='yellow')
# A legend summarizes all this work
legend("bottomright", bg="white", 
       pch=c(21, 21, 22), pt.bg=c("red", "pink", "yellow"),
       legend=c("Best", "Range", "True"))

center

Results and conclusions

The diagram above indicates that varying times by half a minute (corresponding
to the rounding interval in public forecasts of sunrise and sunset times)
yields approximately 25km of uncertainty in geographical position, at this
particular time of year. (Note that a degree of latitude is about 111km.)

Readers interested in exploring the uncertainty through the year should find
the R code simple to modify. It is also easy to express the uncertainty
statistically.

Further discusion of matters relating to solar angles can be found in my
upcoming book [3].

Resources

  1. A website for the SkyView project is
    http://emit.phys.ocean.dal.ca/~kelley/skyview.

  2. A website providing the requisite sunrise and sunset times is
    http://www.timeanddate.com/astronomy/canada/halifax.

  3. Dan Kelley, in preparation. Oceanographic Analysis with R. Springer-Verlag.

  4. Source code: 2014-11-10-solar-navigation.R.

To leave a comment for the author, please follow the link and comment on their blog: Dan Kelley Blog/R.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.

Search R-bloggers


Sponsors

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)