Santek, D., R. Dworak, S. Wanzong, K. Winiecki, S. Nebuda, J. García-Pereda, R. Borde, M. Carranza, 2018: Third AMV Intercomparison Study. Proceedings, 14th International Winds Workshop, Jeju, Korea

Abstract

This study is a continuation and update of the previous “2014 AMV Intercomparison study”, presented in the 12th International Winds Workshop in Copenhagen in 2014.
In this continuation, Atmospheric Motion Vectors (AMVs) calculated with Japan Meteorological Agency’s Himawari-8 satellite data are compared, considering two different input datasets with two different image triplets for 21st July 2016. Image data are equivalent to those used by the “International Cloud Working Group (ICWG) Cloud Intercomparison study”, to improve synergies between both studies. The different centers use a prescribed configuration and their own configuration for the AMV production with these datasets.
Six different institutions have participated in the study (CPTEC/INPE, EUMETSAT, JMA, KMA, NOAA and NWCSAF).

Full text

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
Third AMV Intercomparison Study
David Santek1, Richard Dworak1, Steve Wanzong1, Katherine Johnson1, Sharon Nebuda1,
Javier García-Pereda2, Régis Borde3, Manuel Carranza4
1 CIMSS/University of Wisconsin-Madison, 1225 W. Dayton St., Madison, Wisconsin, 53706 USA
2 NWCSAF/AEMET, Leonardo Prieto Castro 8, Madrid, 28040 Spain
3 EUMETSAT, Eumetsat Allee 1, Darmstadt, 64295 Germany
4 GMV INSYEN@EUMETSAT, Eumetsat Allee 1, Darmstadt, 64295 Germany
Abstract
This study is a continuation and update of the previous “2014 AMV Intercomparison study”, presented
in the 12th International Winds Workshop in Copenhagen in 2014.
In this continuation, Atmospheric Motion Vectors (AMVs) calculated with Japan Meteorological Agency’s
Himawari-8 satellite data are compared, considering two different input datasets with two different image
triplets for 21 July 2016. Image data are equivalent to those used by the “International Cloud Working
Group (ICWG) Cloud Intercomparison study”, to improve synergies between both studies. The different
centers use a prescribed configuration and their own configuration for the AMV production with these
datasets.
Six different institutions participated in the study (CPTEC/INPE, EUMETSAT, JMA, KMA, NOAA and
NWCSAF). This paper is a summary of the full “AMV Intercomparison Technical Report”, which can be
found at: http://www.nwcsaf.org/aemetRest/downloadAttachment/5284. The study has been updated in
November 2018 with two new datasets from EUMETSAT and KMA, which correct two issues related
with the “Common Quality Index (QIC)” and the “Height assignment” respectively.
INTRODUCTION
Two “AMV Intercomparison studies” have taken place in the past up to now: Genkova et al. 2008 &
2010, and Santek et al. 2014. The evolution of the AMV algorithms and of the geostationary satellites
since then defined the need for a “Third AMV Intercomparison study” in 2017-2018. Three main goals
are considered for this new study:
1) To verify the advantages of the calculation of AMVs with the new generation of geostationary
satellites, started with Himawari-8, with better spatial and temporal resolution and new spectral
channels, with respect to those calculated with MSG series.
2) To extract conclusions about the best options for the calculation of AMVs with this new generation
of geostationary satellites, considering the options taken by the different centers for their AMV
calculation.
3) To compute a “Common Quality Index (QIC)” for all centers, to verify if there is a better agreement
between the different AMV datasets.
The report analyzes the AMV algorithms provided by the following six AMV producers. The three-letter
abbreviations are used as identifiers of the AMV datasets throughout the remainder of this report:
BRZ: Brazil Weather Forecast and Climatic Studies Center (CPTEC/INPE)
EUM: European Organization for the Exploitation of Meteorological Satellites (EUMETSAT)
JMA: Japan Meteorological Agency
KMA: Korea Meteorological Administration
NOA: Unites States National Oceanic and Atmospheric Administration (NOAA)
NWC: Satellite Application Facility on Support to Nowcasting (NWCSAF)
China Meteorological Administration (CMA) participated in the previous intercomparison study but not
in this one.

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
The AMV outputs were originated considering two triplets of Himawari-8/AHI infrared (10.4 µm) full–disk
images for 21 July 2016 at 0530-0550 and 1200-1220 UTC, one of which is shown in Figure 1.
Additionally, ECMWF ERA-INTERIM NWP analysis for the given day, for 37 vertical levels every 6
hours, and corresponding cloud products derived by NOAA/NESDIS for the given slots, were provided
for the AMV calculation.
The AMV outputs provided by each AMV algorithm, are analyzed in three independent experiments,
designed to measure differences related to specific aspects of the algorithms. Scripts used in the two
previous intercomparison studies (Genkova et al. 2008 & 2010, and Santek et al. 2014) have been used
again, so allowing for the comparison of the results in the different studies.
Figure 1: Himawari-8 10.4 µm satellite image for 21 July 2016 at 1200 UTC
Each center’s output for the experiments included data for identical variables, as shown in Table 1, with
the exception of BRZ, who did not report the “Quality Index with forecast (QIF)”.
Parameter
Code
Description
1
IDN
Identification number
2
LAT[DEG]
Latitude
3
LON[DEG]
Longitude
4
TS[PIX]
Target box size
5
SS[PIX]
Search box size
6
SPD[MPS]
AMV speed
7
DIR[DEG]
AMV direction
8
PRES[HPA]
AMV pressure
9
L
Low level correction flag
10
NWPSPD[MPS]
Background guess wind speed
11
NWPDIR[DEG]
Background guess wind direction
12
ALB[%]
Albedo
13
CORR[%]
Correlation
14
T[K]
Brightness temperature
15
PRESERR[HPA]
AMV pressure error
16
H
Height assignment method flag
17
QINF[%]
Quality Index without forecast
18
QIF[%]
Quality Index with forecast
19
QIC[%]
Common Quality Index
Table 1: Reported Variables

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
EXPERIMENT 1
In this case, AMV producers extracted cloudy AMVs with the triplet 1200-1220UTC,
using their best options for the AMV calculation, but considering a prescribed target box size, search
scene size and target locations. All AMV extraction processes could be compared this way (tracer
selection, tracer tracking, height assignment and quality control), comparing equivalent AMV datasets.
Figure 2 shows the distribution of parameters (Common Quality Index, speed, direction and pressure)
for the different AMV datasets, with a QIC threshold of 50%:
Figure 2: Distribution of parameters for Experiment 1 considering QIC >= 50% (AMV distribution, Common Quality Index,
Speed, Direction, Pressure) for the different AMV datasets (BRZ: upper left; EUM: lower left; JMA: upper center; KMA:
lower center; NOA: upper right; NWC: lower right).
The distributions look similar for all centers, with the following items to be taken into account:
1) The distribution of direction values for BRZ shows some directions more frequent than other ones;
2) The distribution of the QIC values looks basically similar for all centers;
3) The distribution of AMV pressures is instead very different for the different centers due to the very
different calculation methods - only EUM & NWC being similar because of both using “CCC method”
for the height assignment. This last result is even more evident looking to Figure 3, in which the
scatter plot of AMV pressures of all centers is shown using the EUM pressure as reference.
Figure 3: Scatterplot of collocated AMV pressures for Experiment 1 considering QIC >= 50%, for each center versus EUM
AMV pressures (BRZ in green; JMA in yellow; NOA in black; KMA in red; NWC in pink).

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
When the AMVs are compared to radiosonde winds (in Table 2 using the threshold of 50%, and in Table
3 using the threshold of 80% for the QIC, the best results are for JMA (with a vector RMS of 5 m/s), and
then for NWC and NOA (with a vector RMS of 6-8 m/s). BRZ and EUM show bad results for the low
quality threshold, while much better for the high quality threshold, for which there is more homogeneity
between centers. In addition, there are important differences in the number of AMVs for the different
centers with the prescribed configuration (although in all cases the number of AMVs is larger than in the
previous study with MSG satellite).
N
Pre Bias
Pre RMS
Spd Bias
Spd RMS
Dir Bias
Vec RMS
BRZ
774
0.90
14.68
.1,28
10.00
-13.13
12.61
EUM
473
-1.63
16.52
-1.74
7.86
8.69
12.67
JMA
400
0.50
13.81
-0.91
3.95
1.30
5.74
KMA
859
-0.55
15.16
-1.88
7.61
5.46
10.10
NOA
512
-1.05
14.15
-0.86
6.29
1.10
8.16
NWC
163
-0.69
15.11
-1.14
4.99
-1.47
6.80
Table 2: Experiment 1: Comparison of AMVs (with QIC >= 50%) to radiosonde winds within 150 km. N = number of
matches; Pre Bias = pressure bias; Pre RMS = pressure RMS; Spd Bias = wind speed bias; Spd RMS = wind speed RMS;
Dir Bias = wind direction bias; Vec RMS = vector RMS. The extreme for each category is highlighted: Yellow = worst
value; cyan = best value
N
Pre Bias
Pre RMS
Spd Bias
Spd RMS
Dir Bias
Vec RMS
BRZ
448
0.02
13.97
0.31
6.07
-14.61
8.62
EUM
312
-0.83
16.54
-1.79
6.54
8.31
8.56
JMA
344
0.78
13.77
-1.07
4.07
1.09
5.93
KMA
666
-0.79
15.51
-1.56
6.42
2.78
8.97
NOA
427
-1.49
13.96
-0.89
5.42
0.45
7.52
NWC
132
-0.41
15.08
-0.97
5.01
-5.98
6.94
Table 3: Experiment 1: Comparison of AMVs (with QIC>= 80%) to radiosonde winds within 150 km. N = number of
matches; Pre Bias = pressure bias; Pre RMS = pressure RMS; Spd Bias = wind speed bias; Spd RMS = wind speed RMS;
Dir Bias = wind direction bias; Vec RMS = vector RMS. The extreme for each category is highlighted: Yellow = worst
value; cyan = best value
Considering the comparison of collocated AMVs against the NWP analysis winds, in Table 4 using the
threshold of 80% for the QINF, and in Table 5 using the threshold of 80% for the QIC, the differences
between centers are smaller (with only BRZ really over), and even smaller using the QIC for the filtering.
N
BFN
VD
RMS
VDABF
RMSABF
BRZ
4930
1191
5.90
8.47
5.27
8.16
EUM
4930
1625
3.96
4.93
3.19
4.28
JMA
4930
1793
2.48
2.96
2.24
2.77
KMA
4930
1732
3.69
4.61
2.85
3.79
NOA
4930
1757
3.45
4.30
2.76
3.73
NWC
4930
1763
3.95
4.70
3.09
3.95
Table 4: Experiment 1: Comparison of collocated AMVs (with QINF >= 80%) to NWP analysis winds. N = number of AMVs;
BFN = number of AMVs with Best fit pressure; VD = Vector difference for all AMVs; RMS = Root mean square error for
all AMVs; VDABF = Vector difference for AMVs with Best fit pressure; RMSABF = Root mean square error for AMVs with
Best fit pressure. The extreme for each category is highlighted: Yellow = worst value; cyan = best value
N
BFN
VD
RMS
VDABF
RMSABF
BRZ
8076
2122
5.54
7.53
4.85
7.13
EUM
8076
2655
4.04
4.97
3.24
4.28
JMA
8076
2860
2.59
3.10
2.33
2.89
KMA
8076
2802
3.80
4.73
2.97
3.94
NOA
8076
2854
3.54
4.36
2.82
3.74
NWC
8076
2791
3.99
4.74
3.17
4.03
Table 5: Experiment 1: Comparison of collocated AMVs (with QIC >= 80%) to NWP analysis winds. N = number of AMVs;
BFN = number of AMVs with Best fit pressure; VD = Vector difference for all AMVs; RMS = Root mean square error for
all AMVs; VDABF = Vector difference for AMVs with Best fit pressure; RMSABF = Root mean square error for AMVs with
Best fit pressure. The extreme for each category is highlighted: Yellow = worst value; cyan = best value

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
Considering in Figure 4 the AMV level against the AMV best fit level, it is also clear that JMA AMVs are
near the best fit level; much more than for all other datasets. The maps in the lower panels of Figure 4
depict the best fit displacements above (red) and below (blue) the AMV level, which tend to be in similar
locations for all centers for collocated AMVs.
Figure 4: Experiment 1: Histogram and maps of “AMV best fit pressure – original AMV pressure” for BRZ, EUM, JMA,
KMA, NOA, NWC (from left to right). In the maps, red shows the best fit level is at a higher level; blue shows the best fit
level is at a lower level
EXPERIMENT 2
In this case, AMV producers extracted cloudy AMVs with the triplet 1200-1220UTC,
using their best options for the AMV calculation, and considering their own configuration for target box
size, search scene size and target locations. The differences of each AMV extraction process (with
respect to the previous prescribed configuration) can be compared this way.
Figure 5 shows the distribution of parameters (Common Quality Index, speed, direction and pressure)
for the different AMV datasets, with a QIC threshold of 50%:
Figure 5: Distribution of parameters for Experiment 1 considering QIC >= 50% (AMV distribution, Common quality index,
Speed, Direction, Pressure) for the different AMV datasets (BRZ: upper left; EUM: lower left; JMA: upper center; KMA:
lower center; NOA: upper right; NWC: lower right).

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
The parameter distributions are very similar to those in Experiment 1. With this, the differences in the
height assignment process drive the majority of differences observed. Again, the distribution of the QIC
values looks similar for all centers.
The scatter plot of AMV pressures of all centers using the EUM pressure as reference is shown again
in Figure 6, with equivalent results to those in Experiment 1 (as expected).
Figure 6: Scatterplot of collocated AMV pressures for Experiment 2 considering QIC >= 50%, for each center versus EUM
AMV pressures (BRZ in green; JMA in yellow; NOA in black; KMA in red; NWC in pink).
When the AMVs are compared to radiosonde winds (in Table 6 using the threshold of 50%, and in Table
7 using the threshold of 80% for the QIC, the best results are again for JMA (with a vector RMS of 6
m/s), and then for NOA and NWC (with a vector RMS of 7 m/s). EUM results are much better in
Experiment 2, using their own configuration. The number of AMVs in Experiment 2 with respect to
Experiment 1 changes only significantly for NWC, for which there is an increase of around 15 times.
N
Pre Bias
Pre RMS
Spd Bias
Spd RMS
Dir Bias
Vec RMS
BRZ
942
1.46
14.44
-2.69
11.65
-9.48
15.22
EUM
508
-0.62
15.92
-2.17
7.03
10.08
8.87
JMA
313
-2.94
18.33
-1.36
4.64
-0.83
6.34
KMA
797
-0.64
14.80
-1.55
7.78
-1.41
10.03
NOA
691
-1.58
13.93
-0.90
5.44
1.89
7.62
NWC
2204
-1.02
16.30
-2.17
6.03
0.40
7.85
Table 6: Experiment 2: Comparison of AMVs (with QIC >= 50%) to radiosonde winds within 150 km. N = number of
matches; Pre Bias = pressure bias; Pre RMS = pressure RMS; Spd Bias = wind speed bias; Spd RMS = wind speed RMS;
Dir Bias = wind direction bias; Vec RMS = vector RMS. The extreme for each category is highlighted: Yellow = worst
value; cyan = best value
N
Pre Bias
Pre RMS
Spd Bias
Spd RMS
Dir Bias
Vec RMS
BRZ
619
1.16
13.44
-0.40
7.36
-14.65
9.80
EUM
366
-0.66
14.74
-2.20
6.15
8.43
8.05
JMA
270
-3.43
18.67
-1.40
4.64
-0.83
6.42
KMA
628
-0.84
14.30
-1.21
7.39
-2.66
9.49
NOA
599
-1.69
13.98
-0.88
5.25
0.39
7.48
NWC
2063
-1.19
16.19
-2.11
5.99
0.79
7.85
Table 7: Experiment 2: Comparison of AMVs (with QIC >= 80%) to radiosonde winds within 150 km. N = number of
matches; Pre Bias = pressure bias; Pre RMS = pressure RMS; Spd Bias = wind speed bias; Spd RMS = wind speed RMS;
Dir Bias = wind direction bias; Vec RMS = vector RMS. The extreme for each category is highlighted: Yellow = worst
value; cyan = best value

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
Considering the comparison of collocated AMVs against the NWP analysis winds, in Table 8 using the
threshold of 80% for the QINF, and in Table 9 using the threshold of 80% for the QIC, the differences
between centers are smaller for collocated AMVs.
N
BFN
VD
RMS
VDABF
RMSABF
BRZ
43281
9814
5.60
8.20
5.01
7.90
EUM
43281
13270
3.84
4.96
3.05
4.24
JMA
43281
14572
2.20
2.71
1.99
2.52
KMA
43281
12709
3.75
5.05
3.08
4.51
NOA
43281
13765
3.41
4.26
2.74
3.64
NWC
43281
13588
3.45
4.13
2.79
3.54
Table 8: Experiment 2: Comparison of collocated AMVs (with QINF >= 80%) to NWP analysis winds. N = number of AMVs;
BFN = number of AMVs with Best fit pressure; VD = Vector difference for all AMVs; RMS = Root mean square error for
all AMVs; VDABF = Vector difference for AMVs with Best fit pressure; RMSABF = Root mean square error for AMVs with
Best fit pressure. The extreme for each category is highlighted: Yellow = worst value; cyan = best value
N
BFN
VD
RMS
VDABF
RMSABF
BRZ
56515
13075
5.73
8.35
5.11
8.02
EUM
56515
17533
4.00
5.17
3.17
4.43
JMA
56515
19208
2.27
2.80
2.06
2.62
KMA
56515
16635
3.92
5.25
3.23
4.72
NOA
56515
18163
3.53
4.42
2.84
3.80
NWC
56515
17860
3.55
4.24
2.87
3.65
Table 9: Experiment 2: Comparison of collocated AMVs (with QIC >= 80%) to NWP analysis winds. N = number of AMVs;
BFN = number of AMVs with Best fit pressure; VD = Vector difference for all AMVs; RMS = Root mean square error for
all AMVs; VDABF = Vector difference for AMVs with Best fit pressure; RMSABF = Root mean square error for AMVs with
Best fit pressure. The extreme for each category is highlighted: Yellow = worst value; cyan = best value
Considering in Figure 7 the AMV level against the AMV best fit level, it is again clear that JMA AMVs
are near the best fit level; much more than for all other datasets. Best fit displacements above and
below, tend to be again in similar locations for all centers for collocated AMVs.
Figure 7: Experiment 2: Histogram and maps of “AMV best fit pressure – original AMV pressure” for BRZ, EUM, JMA,
KMA, NOA, NWC (from left to right). In the maps, red shows the best fit level is at a higher level; blue shows the best fit
level is at a lower level

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
EXPERIMENT 3
In this case, AMV producers extract IR10.4 μm cloudy AMVs with the triplet 0530-0550UTC, using their
best options for AMV calculation, and considering their own configurations for target box size, search
scene size and target location (as in Experiment 2). This dataset is used for validation against NASA’s
CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation), which provides an
independent measurement of cloud top heights.
CALIPSO is a line-of-site measurement, so there are few collocations with AMVs (tens of matches only).
Therefore, this evaluation is qualitative as illustrated in the following Figure 8. AMVs are generally near
the cloud base for high-level and semitransparent clouds, and near the cloud top for low- and mid-level
clouds. AMV heights for the different centers are in good agreement in this specific example, in apparent
disagreement with the previous AMV pressure scatter plots.
Figures 8: Experiment 3: Collocation of AMVs (defined as black asterisks *), with CALIPSO cloud measurements for BRZ
(upper row left), JMA (upper row right), NOA (center row left), EUM (center row right), KMA (lower row left) and NWC
(lower row right).

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
CONCLUSIONS
In general, the differences in AMV datasets for each centre between Experiment 1 and 2 are basically
related to the number of AMVs. In addition, the differences in AMV datasets for different centres are
much more related to the height assignment process than to the use of a prescribed or a specific
configuration.
Another important conclusion is that the distribution of the Common Quality Index values is very similar
for all centers, and the use of the QIC has a real skill in filtering collocated AMVs for an improved
statistical agreement.
Considering also specific conclusions for the different centers:
BRZ - Brazil Weather Forecast and Climatic Studies Center (CPTEC/INPE)
The performance of BRZ algorithm has improved with respect to the previous AMV intercomparison,
with better agreement with other centers (especially, for high Quality index thresholds and collocated
AMV data). Anyhow, there still exists room for improvement: large differences in the AMV pressures,
and the need to verify the direction histograms, with some directions much more frequent than other
ones.
KMA - Korea Meteorological Administration
The KMA algorithm performed similarly to the results from the previous AMV intercomparison. Overall,
the comparisons to rawinsondes and model background were in the middle of the distributions. KMA
algorithm is reasonably good, but it needs still to define its final stable version.
NOA – United States National Oceanic and Atmospheric Administration (NOAA)
NOAA agreement compared to other centers improves over the previous study. NOAA algorithm has
now the second best statistics, along with NWCSAF. An element for analysis is the vertical distribution
of AMVs, with no AMVs present between 450-700 hPa (in contrast to other algorithms).
NWC – Satellite application facility on support to Nowcasting (NWCSAF)
NWCSAF algorithm has the second best statistics, along with NOAA. The algorithm is basically similar
to the one in the previous study, and due to this stability, the performance is similar to the one found
then. An element for analysis is that some directions for Himawari AMVs are more frequent than other
ones in the vicinity of 90 degrees.
EUM - European Organization for the Exploitation of Meteorological Satellites (EUMETSAT)
The behavior of EUMETSAT algorithm is much better when the Quality index threshold is high (80%)
and the specific configuration is used. In these circumstances, the performance is basically similar to
that of other centers. The similarity in the height assignment with NWC center is also to be noticed, due
to both using “CCC method”.
JMA - Japan Meteorological Agency
JMA algorithm has the best overall performance considering all validation and checking elements, most
likely due to its updated height assignment procedure: “optimal estimation method using observed
radiance and NWP vertical profile”. This is the most important change in all AMV algorithms since the
previous AMV intercomparison. However, it is to be studied if the small difference between the AMVs
and the background NWP has a good impact in later applications, like NWP assimilation.

Proceedings for the 14th International Winds Workshop
23-27 April 2018, Jeju City, South Korea
ACKNOWLEDGEMENTS
The International Winds Working Group (IWWG) wants to thank EUMETSAT, the Satellite Application
Facility on support to Nowcasting (NWCSAF) and Agencia Estatal de Meteorología (AEMET) for the
funding and support of this project.
The NWCSAF and the IWWG also want to thank colleagues from CIMSS/University of Wisconsin-
Madison for the work done (a 208 page report) with a very tight schedule.
They also want to thank colleagues in the different AMV production centers for the effort to provide the
AMV datasets for the AMV Intercomparison: Renato Galante Negri (BRZ), Régis Borde and Manuel
Carranza (EUM), Javier García Pereda (NWC), Oh Soomin and Lee Byungil (KMA), Kenichi Nonaka
and Kazuki Shimoji (JMA) and Wayne Bresky (NOA).
REFERENCES
Genkova, I., R. Borde, J. Schmetz, J. Daniels, C. Velden, K. Holmlund, 2008: “Global atmospheric
motion vectors intercomparison study”, 9th International Winds Workshop, Annapolis, Maryland, USA,
April 2008. Available online at:
https://www.eumetsat.int/website/wcm/idc/idcplg?IdcService=GET_FILE&dDocName=PDF_CONF_P5
1_S4_20_GENKOVA_V&RevisionSelectionMethod=LatestReleased&Rendition=Web
Genkova, I., R. Borde, J. Schmetz, C. Velden, K. Holmlund, N. Bormann, P. Bauer, 2010: “Global
atmospheric motion vector intercomparison study”, 10th International Winds Workshop, Tokyo, Japan,
February 2010. Available online at:
https://www.eumetsat.int/website/wcm/idc/idcplg?IdcService=GET_FILE&dDocName=PDF_CONF_P5
6_S4_02_GENKOVA_V&RevisionSelectionMethod=LatestReleased&Rendition=Web
Santek, D., J. García-Pereda, C. Velden, I. Genkova, S. Wanzong, D. Stettner, M. Mindock, 2014: “2014
AMV Intercomparison Study”, 12th International Winds Workshop, Copenhagen, Denmark, June 2014.
Technical Report available online at: http://www.nwcsaf.org/aemetRest/downloadAttachment/225;
Summary available online at: http://www.nwcsaf.org/aemetRest/downloadAttachment/226.
Santek, D., R. Dworak, S. Wanzong, K. Winiecki, S. Nebuda, J. García-Pereda, R. Borde, M. Carranza,
2018: “Third AMV Intercomparison Study”, 14th International Winds Workshop, Jeju City, South Korea,
Updated in November 2018.
Technical Report available online at: http://www.nwcsaf.org/aemetRest/downloadAttachment/5284.