More static maps of path
                      IOTA/IOTA-ES occultation update for                      
             (433) Eros / HIP 33731 event on 2019 Mar 12, 05:48 UT             
                           Visible from W USA, Mexico                          


Summary
-------
On 2019 Mar 12 UT, the 16.4 km diameter asteroid (433) Eros will occult a 6.7
mag star in the constellation Monoceros for observers along a path across W
USA, Mexico.

In the case of an occultation, the combined light of the asteroid and the star
will drop by  3.86 mag to 10.53 mag (the magnitude of the asteroid) for at most
2.0 seconds.

This update is based on UNSO/Flagstaff astrometry for the asteroid kindly
provided by Hugh Harris, astrometry for the asteroid kindly provided by Bill
Owen, astrometry for the asteroid kindly provided by the IAU Minor Planet
Center.

This work has made use of data from the European Space Agency (ESA)
mission Gaia (http://www.cosmos.esa.int/gaia), processed by the
Gaia Data Processing and Analysis Consortium
(DPAC, http://www.cosmos.esa.int/web/gaia/dpac/consortium).
Funding for the DPAC has been provided by national institutions, in
particular the institutions participating in the Gaia Multilateral Agreement.

The event at a glimpse
----------------------
    * Rank: 99
    * date and approx. time of event: 2019 Mar 12, 05:39 - 2019 Mar 12, 05:58
    * geocentric midpoint of event [JD]: 2458554.74236667
    * magnitude of target star: 6.70
    * magnitude drop [mag]: 3.86
    * estimated maximum duration [s]: 2.0
    * Moon: 28 % sunlit, 52į distance
    * Sun: 113į distance
    * rough path description: W USA, Mexico

The occultation path
--------------------
    * approximate projected width [km]: 23
    * 1 sigma uncertainty interval [path widths]: +/- 0.18
    * 1 sigma uncertainty interval [seconds]: +/- 0.5
    * 1 sigma uncertainty interval approx RA,DE ["]: (+/- 0.015, +/- 0.014)
    * 1 sigma uncertainty ellipse (major, minor, PA): (0.017", 0.012", 128į)
    * approx speed of asteroid's shadow [km/s]: 8.1670
    * website for maps:
          http://www.asteroidoccultation.com

--------------------------
Derek Breit's link
JPL prediction path coordinates
Jon Giorgini wrote 2019 Feb. 15:

Outside of published papers, we give 3-sigma pointing uncertainties.

People usually take uncertainties to be absolute error bounds and 99.7% is closer to this habit than 1-sigma values. It can also cover small uncalibrated error sources.

There is a new solution #656 in Horizons that includes NEAR s/c data
as a type of synthetic radar observable.

If interested in details, they are below.

-------------------------------------------------------------------------------
Jon Giorgini                       |  Navigation & Mission Design Section
Senior Analyst                     |  Solar System Dynamics Group
Jon.D.Giorgini@jpl.nasa.gov        |  Jet Propulsion Laboratory
-------------------------------------------------------------------------------

#1) The solution #656 now in Horizons include three occultation
measurements from 2011 and one from 1975. These were always included,
but weights have been tightened to 0.1" for the 2011 data:

                                   Residual
                          norm    RA     DEC  std_error
E20111213.451189 244 x   0.508  0.007 -0.052 0.10M0.10M    095451.675+372456.73
E20111124.421842 244 x   0.253 -0.025  0.008 0.10M0.10M    085629.307+441505.99
E20111023.345090 244 r   0.484 -0.010  0.047 0.10M0.10M    065303.436+464312.47
E19750124.018890 244     1.659 -0.321  0.086 0.2 r0.2 r    074423.906+242359.87

The 1975 occultation measurement is weighted at 0.2", but isn't
fitting to that level. The "norm" column is the residual divided by
the assigned measurement uncertainty.

#2) Unclear what NEAR data Dave had in mind; navigation
solution run-outs from that time terminate in 2002.

However, 54 synthetic delay measurements previously reduced
to Eros COM (derived from DSN radiometric NEAR tracking as part
of the Konopoliv et al gravity solution) have been added.

There are some unresolved minor inconsistencies in that data at
the 10s of meters level, and then a 1.8-sigma inconsistency
with a single recent radar echo measurement (~2.4 km in range),
so some details yet to be perfected but usable.

Adding the NEAR data to the fit moves the solution about 10 km at
the present time.

Differencing angular predicts with and without the NEAR data ...

     Date__(UT)__HR:MN     R.A.__(airls-apparent)__DEC
656: 2019-Feb-16 00:00     05 52 26.8672 +13 33 05.995
654: 2019-Feb-16 00:00     05 52 26.8669 +13 33 05.942
                           -------------  ------------
                                  0.0003s        0.053"

... which is within the uncertainty bounds quoted
yesterday for s654.

#3) Fitting data back to 1893 seems fine.

Comparison finds that two fits to data-arcs 1963-2019 vs. 1893-2019
(while changing nothing else) move the solution only ~1 km, slightly
less than the formal 1-sigma for the new s656.

However, including the old data (very) marginally improves the fit
residual statistics for the NEAR spacecraft data. Since the
preference is to keep as much data as possible and not arbitrarily
exclude, it was kept.

Below, "short" is for an arc 1963-2019, "long" is for an arc 1893-2019.
Smaller numbers are better.

       Normalized residuals            Delay   Doppler     Total  Radar+Opt+
Sat
short: Mean weighted residual          0.310     0.495     0.585
long : Mean weighted residual          0.307     0.495     0.583

short:  Normalized RMS                 0.416     0.655     0.425       0.285856
long :  Normalized RMS                 0.415     0.655     0.423       0.274229

Insignificantly better is at least a little in the right direction.

#4) Formal 3-sigma uncertainties for s656

Date__(UT)__HR:MN     RA_3sigma DEC_3sigma SMAA_3sig SMIA_3sig    Theta
2019-Feb-16 00:00         0.019      0.010     0.021     0.001  -27.126

In terms of heliocentric ICRF vector position 1-sigma uncertainty:

2019-Feb-16 00:00:00.0000 TDB, units are km:
                    X 1-sig                Y 1-sig                  Z 1-sig
sigmas:         1.21045400E+00          9.64330085E-01          8.62131980E-01

Added DWD 2019 Mar 08: So 1-sigma can be 1.2 km in the sky plane, which compares with 2.9 km 1 sigma for Steve P.ís path in the sky plane. This prediction is 4.44 km north (offset -4.44 km) at http://www.poyntsource.com/New/Google/20190312_62652.HTM .

-------------------------------------------------------------------------------
Jon Giorgini                       |  Navigation & Mission Design Section
Senior Analyst                     |  Solar System Dynamics Group
Jon.D.Giorgini@jpl.nasa.gov        |  Jet Propulsion Laboratory
-------------------------------------------------------------------------------

>From
dunham@starpower.net  Thu Feb 14 16:24:23 2019
>Date:
Thu, 14 Feb 2019 19:24:14 -0500 (EST)
>From:
dunham@starpower.net
>To: Jon D Giorgini <
Jon.D.Giorgini@jpl.nasa.gov>
>Cc: ray sterner <
ray.sterner@jhuapl.edu>, davegault@bigpond.com,
>       
stevepr@netstevepr.com, D Herald <D.Herald@bigpond.com>
>Subject: Re: Occultations by (433) Eros
>
>Jon,
>
>Thanks for this, but at first, I was a little surprised by the errors you quote, significantly larger than the ones Steve Preston obtained for another occultation by Eros that will occur
tomorrow night, Feb. 16 UT (that path also crosses Arizona, another 11 th -mag. star, but again, it looks like it will be overcast over the entire state). Then I noticed that your errors are 3-sigma, while Steveís are 1-sigma - for the Feb. 16 event, Steve obtained formal errors of 0.028" and 0.016" error ellipsoid with PA 106 deg.; seehttp://www.asteroidoccultation.com/2019_02/0216_433_63298_Summary.txt .So if multiplied by 3, those are not very much smaller than yours. Part of the problem with your solution may be the inclusion of old astrometric data going back to 1893; the earliest data Steve considered start in 1961. Of course, the radar obs. are line-of-sight, and you donít have very many of them. The astrometric positions from the 2011 occultations should help; they are probably already in the MPC observations of Eros, but perhaps not yet with the Gaia DR2 data for the stars; if re-reduced with the Eros shape model and Gaia DR2 data, the 2011 occultation points should have very high weight. Iím not sure how the NEAR data are factored into either your or Steveís solutions - possibly through the MPC?
>
>Also, asking the others, how can we get the Near/Eros shape model in a form where we can display its projection at the time of the
March 12 th occultation, like we can do for other asteroids? I suppose we can do that at one of the shape model sites I gave in my earlier message (that is, give an Earth-perspective view of Eros at the occultation epoch), just havenít looked into that yet. Since itís just one asteroid, it is probably easier to use some other online resource to obtain the projections manually, then to try to put the Eros shape model in a form that can be used by our current occultation prediction tools.
>
>David

------------------------------
Prediction file based on the JPL Horizons orbital elements, for use in Occult 4.6 (save link as)
Some discussion of the event at the bottom of this page