Simulate targets for a specific sky area (and get expected rates)

Simulate targets for a specific sky area (and get expected rates)#

Sometime, you do not want to simulate target for the entire sky but only for a specific part of the sky. We call that a skyarea. This skyarea is supposed to be a shapely.(Multi)Polygon and can be directly provided in the draw() (or from_draw()) method of a transient. As such, it will be passed to any fucntion that have skyarea as their parameters. This is notably the case of skysurvey.tools.random_radec() that is communly used to get the simulated target coordinates.

This is particularly useful when dealing with deep-field surveys that only observe specific fields in the sky.

Load a mock GridSurvey having 5 deep-fields#

import skysurvey
from skysurvey import examples

survey = examples.get_mock_gridsurvey()

fig = survey.show(autoscale=True)
../_images/757c5eb834395e77036b81a2c1bbf9073957c31378a7d1ff2360467fa73b96c2.png

.get_skyarea()#

GridSurveys have a get_skyarea() method that returns a (Multi)Polygon corresponding to each of its observing field. This is a shapely.geometry

skyarea = survey.get_skyarea()
skyarea
../_images/a975c50bfa2beae2861058694c32be814032d339eb33919bbfd9b4ac9b4907a2.svg

.get_timerange()#

Survey also have a .get_timerange() method that provide the fist and last observation.

tstart, tstop = survey.get_timerange()

.from_draw(skyarea=, tstart=, tstop= )#

Hence, you can simply load a transient using these three keywords

targets = skysurvey.SNeIa.from_draw(tstart=tstart, tstop=tstop,
                                    skyarea=skyarea, 
                                    zmax=1)

from cartopy import crs
origin = 180
fig = survey.show(autoscale=True, origin=origin)
# Show overlapping targets, we use the crs.PlateCarree() for the non-healpy survey.show()
ax = fig.axes[0]
ax.scatter(targets.data["ra"].values-origin, targets.data["dec"].values, 
          transform=crs.PlateCarree(), # Important !
          alpha=0.9, s=3, lw=0)

<matplotlib.collections.PathCollection at 0x2a1c1ff10>
../_images/612b760264873f7f8ce79e9e0ebc01d2b9b8b565e220152cb32da43354c3ac96.png

skyarea.buffer()#

Since skyarea is a shapely.(Multi)Polygon, it has a buffer method, if you want to simulate targets around the fields.

Let’s say we want to simulate 2 degrees around the fields and not just overlapping the fields.

simply do:

targets = skysurvey.SNeIa.from_draw(tstart=tstart, tstop=tstop,
                                    skyarea=skyarea.buffer(2), 
                                    zmax=1)

from cartopy import crs
origin = 180
fig = survey.show(autoscale=True, origin=origin)
# Show overlapping targets, we use the crs.PlateCarree() for the non-healpy survey.show()
ax = fig.axes[0]
ax.scatter(targets.data["ra"].values-origin, targets.data["dec"].values, 
          transform=crs.PlateCarree(), # Important !
          alpha=0.9, s=3, lw=0)

<matplotlib.collections.PathCollection at 0x2a1b8c580>
../_images/336dc02b94e660aa2152d069a5fe898a5112523e4520b634a660da2932aa0c14.png

No size provided, how come ?#

Transients like SNeIa have a pre-defined rate (see targets.rate), hence, given tstart and tstop, the draw() method derives nyears. Then, given the inputs skyarea and zmax, it computes size.

tip: you can combine tstop + nyears (-> tstart) | tstart + nyears (->tstop). If you provide both nyears and size, nyears will be ignored for the size computation.

tip: using skyarea and nyears, you can

targets.rate # volumetric rate, in Gyr^3/year

23500.0

Using skyarea and nyears to get a quick look at how many targets should exist#

You can combine skyarea and nyears to have a quick look at how many targets are supposed to be created by nature up to a given redshift (or redshift shell).

Say you want to know how many targets nature should create, in 3 years, for a 9deg^2 fields between \(z\in[0.3,0.6]\)

from shapely import geometry
skyarea = geometry.box(0, 0, 3, 3)
print(skyarea.area)
skyarea

9.0
../_images/4eb936e1c2567d0adc052933295db557b5933caa97a45e4fe663f557bc8d5e06.svg 
snia= skysurvey.SNeIa.from_draw(nyears=3, skyarea=skyarea, zmin=0.3, zmax=0.6)
len(snia.data)

635

So: 635 SNeIa !