3rd ACS Workshop
and advanced course
ESO Garching Headquarter,
January 15-19, 2006
Atmospheric Transmission at Microwaves (ATM):
ALMA software on atmospheric effects and their correction
Juan R. Pardo1
Col.: J. Cernicharo1, E. Serabyn2, , F. Viallefond3, M. Wiedner4, R. Hills5, J. Richer5
Consejo superior de Investigaciones Científicas (Spain), (2) California Institute of Technology (USA), (3) LERMAObservatoire de Paris, (4) Köln University (Germany), (5) Cavendish Laboratory (UK),
1. Atmospheric mm/submm refractivity
a. Imaginary Part (absorption spectrum)
b. Real Part (Phase delay terms)
c. Scattering by hydrometeors
2. The atmospheric problem for ALMA
3. ATM: Atmospheric software for ALMA
(1)
1. Atmospheric mm/submm Refractivity
1.0 + Gas-phase contribution + Hydrometeros contribution
Impact approximation
collision << 1 / 
Van-Vleck-Weisskopf line profile
1c. Scattering. Polarization must be included
1a. Imaginary
Part (absorption)
1b. Real Part
(phase delay)
I=
( )
I
Q
« Intensity vector »
K: 2x2 matriz de extinción : vector de emisión
S: 2x2 scattering matrix
K=2S+ 
Chajnantor zenith transmission for 0.5 mm H2O / Water lines / Oxygen lines / ozone lines / H2O-foreign / N2-N2 + N2-O2 + O2-O2
1a. Atmospheric absorption
• "Line" spectrum: rotational transitions of H2O, O2, O3,
N2O, CO, and other trace gases.
• "Hydrometeor" contribution: absorption & scattering
by liquid, ice and other particles.
• Collision-induced absorption by N2-N2, O2-O2 and N2O2, H2O-N2, H2O-O2 mechanisms
1a. Atmospheric absorption: Direct measurements with
FTS experiments at Mauna Kea, Chajnantor & Sout Pole
• Accurately measure the shape of the terrestrial longwave
spectrum.
• Solve the « excess of continuum » problem.
• Input to build an state-of-the-art absorption models:
ARTS (Bremen University) AM (CfA, Harvard) ATM (CalTech, CSIC)
•
•
•
•
Schematics of FTS experiment
Characteristics of CSO-FTS
Mounted on Cassegrain focus of
telescope for dedicated obs. runs.
Detector: 3He cooled Bolometer
Moving arm: 50 cm ~ 200 MHz
resolution
Filters: 7 different (165 to 1600 GHz)
CSO-FTS: Some of the best data sets, subsequently used to study the collision-induced non-resonant absorption
a
Zenith Transmission
ALMA overlap
Frecuency (GHz)
The collision-induced non-resonant absorption
Serabyn, Weisstein, Lis & Pardo,
Appl. Optics, 37:2185, 1998
Matsushita, Matsuo, Pardo, &
Radford, PASJ, 51:603, 1999.
Pardo, Serabyn & Cernicharo,
JQSRT, 68:419, 2001
a
Pardo, Cernicharo & Serabyn, IEEE
TAP, 49:1683, 2001
Pardo, Wiedner, Serabyn, Wilson,
Cunningham, Hills & Cernicharo, Ap.
J. Suppl., 153:363, 2004
Pardo, Serabyn, Wiedner &
Cernicharo, JQSRT, 96:537, 2005
Absorption lines taken out in
the analysis. Remaining
absorption due to collisioninduced non-resonant
mechanisms
Very important result for
submm astronomy
Resulting model
Example 1: Opacity terms in wet atmospheric
conditions at Mauna Kea
Example 2: Atmospheric transmission at high
frequencies in dry Mauna Kea conditions
1.0
Transmission
0.8
0.6
0.4
0.2
0.0
Frequency (GHz)
Model (0.5 mm H2O)
CSO data 3/98
CSO data 7/99
Atacama data 6/98
1. Atmospheric mm/submm Refractivity
1.0 + Gas-phase contribution + Hydrometeros contribution
Impact approximation
collision << 1 / 
Van-Vleck-Weisskopf line profile
1c. Scattering. Polarization must be included
1a. Imaginary
Part (absorption)
1b. Real Part
(phase delay)
I=
( )
I
Q
« Intensity vector »
K: 2x2 matriz de extinción : vector de emisión
S: 2x2 scattering matrix
K=2S+ 
1b. Atmospheric phase delay (wet component)
1. Atmospheric mm/submm Refractivity
1.0 + Gas-phase contribution + Hydrometeros contribution
Impact approximation
collision << 1 / 
Van-Vleck-Weisskopf line profile
1a. Imaginary
Part (absorption)
1c. Scattering. Polarization must be included
1b. Real Part
(phase delay)
I=
()
I
Q
« Intensity vector »
K: 2x2 matriz de extinción : vector de emisión
S: 2x2 scattering matrix
High cirrus
(ice particles)
K=2S+ 
2. The atmospheric problem for ALMA
atm
(baseline, wind,
atmospheric status)
~ 20 m
a
Atmospheric water vapor fluctuations
gas,i(x,y,z,t)
P(x,y,z,t)
T(x,y,z,t)
Refractive index field
n(x,y,z,t)
t=t0; nd
Flat ground for interferometry
Antennae positions:
xi,yi)  (Rxy)i
: Line of sight
coordinate
Wavefront delay
(x,y)  (Rxy)
(
< (rxy) (Rxy-rxy)>
Radiative transfer along coordinate  at t=t0
Atmospheric brightness temperature
Power spectrum of delay
TB(x,y)  TB(Rxy)
P(qxy)  P(qxy)
< TB(rxy) TB(Rxyrxy)>
AT (t)
A(t)
B
1D FT
P()
PT ()
B
2D FT
Temporal autocorrelation functions
2D FT
Power spectrum of TB
PT (qxy)
Delay autocorrelation
function
A(Rxy)
Frozen screen; y=0 x=vt
TB autocorrelation
function
AT (Rxy)
B
B
All functions are
 dependent
Example of atmospheric water vapor field: Kolmogorov Screen
With an
interferomerter and
a reference point
source
P(qxy)
PTB,()  PH O()  P()
2
at astro
With one antenna
we can measure
TB(t), from which
we obtain:
<TB(t') TB(t-t')>
y (m)
PT ,()
B

Chose
very
sensitive to H2O so
that we track
PH O()
2
With two antennas:
P() at 
astro
Temporal Power Spectra: Px()
The slope informs on the scale of phenomena
5km
500m
50m
d for
v=5 m/s
PTB,() measured on
Nov/25/2001 at Mauna Kea
1 = 183.317.8 GHz
2 = 183.313.0 GHz
3 = 183.311.3 GHz
6km
Outer sc.
0
-5/3
outer
2D
70m
-5/3
2D
-8/3
3D
P()
Instr.
Time 1000s
100s
10s
-8/3
3D Kol
measured on
Nov/25/2001 at Mauna Kea
at astro = 230.5 GHz
Target source: Uranus
Base line: 48 m
PTB,()  PH O()  P() OK / 70 m - 6 km : 2D Kolmogorov screen
2
Phase correction can be performed using water vapor
radiometry at frequencies sensitive to H2O
8.47 μm pathlength / μm PWV
612 μm
10.41 μm/μm
6.65 μm/μm
462 μm
353 μm
Assuming 0.4 K noise in 1 s in all three
channels of the Water Vapor Radiometer
Uncertainty in determining the PWV: 2.3 μm
Taking an average Precipitable Water Vapor
amount of 0.5 mm, we have:
23 mk/μm
60 mk/μm
173 mk/μm
The goal of 15 μm pathlength correction
accuracy for ALMA should be reachable with
this technique in time scales of the order of 1 s.
CURRENT PROTOTYPES ALREADY MEET
THESE REQUIREMENTS
Phase fluctuations: Correction using 183 GHz water vapor radiometers
Observing Frequency: 230.5 GHz, PWV to phase conversion factor: 1.944 deg/m
a
Time since the beginning of the observation (min) on Nov. 25, 2001
3. ATM: Atmospheric software for ALMA
• Starting point: Fortran code developed during 20 years.
• Contract with ESO to develop ATM for alma.
• Fortran library created to encapsulate "fundamental" physics of
the problem (C++).
• C++ interface developed specifically for ALMA needs.
• Updated via CVS.
• Doxygen documented.
• Test examples provided using real FTS and WVR data.
• Work done withing the TelCal working group.
Old ATM fortran code
Ground Temperature
Tropospheric Lapse Rate
Ground Pressure
Rel. humidity (ground)
Water vapor scale height
Primary pressure step
Pressure step factor
Altitude of site
Top of atmospheric profile
ATM_telluric
ATM_st76
ATM_rwat
ABSORPTION COEFS.
kh2o_lines (NPP)
kh2o_cont (NPP)
ko2_lines (NPP)
kdry_cont (NPP)
kabs (NPP)
kminor_gas (NPP)
PHASE FACTORS
total_dispersive_phase (NPP)
total_nondisp_phase (NPP)
Atmospheric radiance
Frequency
RT_telluric
User
Other sources
to obtain
atmospheric
parameters
DATA_xx_lines
DATA_xx_index
INI_singlefreq
Air mass
Bgr. Temp.
INTERFACE LEVEL
1st guess of water vapor column
Layer thickness (NPP)
Number of atmospheric layers, NPP
Layer pressure (NPP)
Layer temperature (NPP)
Layer water vapor (NPP)
Layer_O3 (NPP)
Layer CO (NPP)
Layer N2O (NPP)
xx
all species
ABS_h2o
ABS_o2
ABS_co
ABS_o3_161618
ABS_h2o_v2
ABS_o2_vib
ABS_n2o
ABS_o3_161816
ABS_hdo
ABS_16o17o
ABS_o3
ABS_o3_nu1
ABS_hh17o
ABS_16o18o
ABS_o3_161617 ABS_o3_nu2
ABS_hh18o
ABS_cont
ABS_o3_161716 ABS_o3_nu3
Applications: INV_telluric for WVR and FTS
data (available in current release)
DEEP LEVEL(LIBRARY)
I/O Parameters
ALMA-ATM C++ structure:
Collaboration diagram between the most important classes
Class AtmProfile: Profiles of physical conditions & chemical abundances
Class AbsorptionPhaseProfile: Profiles of refractive index for an array of frequencies
Class WaterVaporRadiometer: WVR system in place for phase correction
Class SkyStatus: Relevant atmospheric information for antenna operations