**!! IMPORTANT INFORMATION !!** As of October 31, 1996 the North/South
flow angle corrections have NOT yet been done for the hourly average files!   
***********************************************************************

Here are some notes about the content of the hourly average subsets.
Please note that these averages are only for interplanetary SOLAR WIND data.

1) Each averaged parameter consists of three fields of information,
i.e. a "triplet".  The first column is the average value, the second is 
the standard deviation (std. dev.) associated with points contributing to 
that average, and the third is the number of individual data points (# pts.) 
used to calculate that average.  Because different parameters are calculated 
using slightly different assumptions, there may be variations in the number 
of points used for the various averages in an hour.  The averages were 
calculated from fine resolution parameters that are based on a non-linear
fit to a convected, isotropic Maxwellian model.

2) These are only averages of our 'best' parameters.  A value of 0.
means that we had no values of that parameter for that hour.  If
you're desperate, we can (after some discussion) give you less
accurate parameters for specific times, but caveat emptor...

3) We use the convention Jan. 1 = DOY 1.  Decimal times are the middle
of each hour.  Please note that the decimal year is double precision,
e.g. 1994.xxxxxxx where xxxxxxx is fraction of year.

4) The velocity coordinates are in GSE, meaning +x toward Sun, +z
towards the North, perpendicular to the (Earth's) ecliptic, y for a
right-hand system (+y in the direction opposite to Earth's motion).
Everything is km/s.  Effects due to Earth's motion around the Sun were
removed from the values before averaging.  It is important to note
that while for each data point, speed = sqrt(Vx*Vx + Vy*Vy + Vz*Vz),
that is NOT true for each hour's average velocity component.  Therefore, 
the average speed may not (is probably not) equal to the square root of the
squares of the components.

5) Thermal speed is the most probable thermal speed (i.e., the square
root of [2kT/m(proton)]). To convert thermal speed to temperature in
eV, multiply 0.0052 by the square of the thermal speed.

6) The angles are in degrees.  Azimuth is E/W, with each meaning 'from
the (E/W),' while elevation is N/S with the same meaning.  For signs,
positive azimuth angle means flow from the W; positive elevation angle
means flow from the S.  If we don't get good angles, you don't get any
velocity components.  Fine resolution data for the speed, velocity
components and angles are not available in those cases.  Again, note
that the aberration in velocity due to Earth's motion around the Sun 
was removed from the data values before averaging. 

7) The FORTRAN format statement used to write this file was:

	write (year,day_of_year,hour,dec_yr,    speed, st.dev., # of points,
		Vx, st.dev., # of points,       Vy, st.dev., # of points,
		Vz, st.dev., # of points,       thermal speed, st.dev., # of points,
		density, st.dev, # of points,   azimuth, st.dev., # of points,
		elevation, st.dev., # of points)

	format (3I4,f16.7,8(3x,3e12.3))

8) For papers and presentations using these data, please acknowledge 
that you received them from the MIT Space Plasma Physics Group. Please
feel free to contact us if you have questions about any parameters.

9) Please send us a copy of papers, presentations, et cetera using
these data.

10) If you have any questions, please contact

   Pamela A. Milligan      pam@space.mit.edu
   Dr. Alan J. Lazarus     ajl@space.mit.edu