 |
| Titel |
The New Version of Brightness Temperature Product of Microwave Radiometer AMSR2 onboard GCOM-W1 Satellite |
| VerfasserIn |
Takashi Maeda, Keiji Imaoka, Marehito Kasahara |
| Konferenz |
EGU General Assembly 2015
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 17 (2015) |
| Datensatznummer |
250113548
|
| Publikation (Nr.) |
 EGU/EGU2015-13763.pdf |
|
|
|
|
|
| Zusammenfassung |
| The Japan Aerospace Exploration Agency (JAXA) launched the GCOM-W1 (Global Change
Observation Mission 1st - Water) satellite on 17 May 2012 (UT). The main mission of the
GCOM-W1 satellite is to monitor the global water cycle, for which it has onboard
the Advanced Microwave Scanning Radiometer-2 (AMSR2). The AMSR2 has
multifrequency microwave receivers, and like many other satellite-borne microwave
radiometers currently in orbit, its multifrequency feedhorns share a single rotating antenna
dish.
Because AMSR2 is designed to make a sharp antenna beam for each frequency, the FOV
(field of view) size of each frequency differs substantially from one another: the higher the
frequency, the smaller the FOV size. Therefore, the multifrequency data (brightness
temperature, TB) represent responses of areas quite different from one another, even if
the multifrequency TB values are obtained by the same measurement. Lack of
consideration of the difference in FOV sizes will be an important reason for the
deterioration of the accuracy of geophysical values retrieved from multifrequency TB
values.
To solve this problem, we can consider to replicate a large FOV of low-target frequency
by adequately summing small FOV of high-source frequency spreading within it. One way to
achieve this is with the Backus-Gilbert (BG) method, and there have been many studies
related to this issue. The BG method provides a way to calculate the intensities for
small-source FOVs to replicate a large-target FOV as accurately as possible. TB
values that correspond to source FOVs (source TB values) are averaged using these
intensities as weighting coefficients. Thus, the TB synthesized from the source
TB values, i.e., the weighted mean of the source TB values becomes equivalent
to the source frequency’s TB measured by a target FOV. The source frequency’s
synthesized TB and the target frequency’s TB enable us to evaluate more accurately the
difference in the microwave emission characteristics between the two frequencies, which
contributes to the improvement of the accuracy of geophysical values retrieved
from them. Depending on the calculation, the weighting coefficients can be positive
or negative. If more negative weighting coefficients appear to replicate a target
FOV accurately, an unexpected value for the synthesized TB is more likely to be
calculated. Therefore, the appearance of negative weighting coefficients should be
controlled while a target FOV is replicated accurately. We investigated the most
appropriate implementation of the BG method for the L1R product, and the smoothing
factors used in the method were dynamically determined for all modified brightness
temperatures.
JAXA released two data products for the TB of the AMSR2: level 1B (L1B) and level 1R
(L1R). The L1B product includes TB values from the native FOV of each channel, and the
L1R product includes the TB values modified with weighting coefficients using the BG
method. These products are available to free download from the GCOM-W1 Data Providing
Service (https://gcom-w1.jaxa.jp/), and updated to the new version (2.0) in April 2014. In this
paper, we present the details of the new version of AMSR2’s L1B and L1R products. |
|
|
|
|
|
|