 |
| Titel |
Deriving a relationship between the radiative power and the SWIR radiance for Gas Flares |
| VerfasserIn |
Alexandre Caseiro, Johannes W. Kaiser, Angelika Heil, Gernot Rücker, Joachim Tiemann, David Leimbach |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250129106
|
| Publikation (Nr.) |
 EGU/EGU2016-9176.pdf |
|
|
|
|
|
| Zusammenfassung |
| Flaring occurs in many regions and is a source of black carbon (BC) globally, among other
pollutants. At higher latitudes, flaring is the main source of BC which, upon deposition on
snow, lowers its albedo. Therefore, knowing the location and emissions of flares would be a
valuable input to climate models.
The main goal of this work is to derive a working relationship between Shortwave
Infrared at 1.6 μm (SWIR) radiance and Radiative Power (RP). Such a relationship will be
helpful in the processing chain use to determine the emissions from gas flares for upcoming
instruments, such as the SLSTR on Sentinel-3.
In order to derive the relationship between RP and SWIR, RP assigned to Gas Flares
(GFs) observed by the small German satellite BIRD is related to the SWIR radiance time
series as observed by AATSR at that location in a temporal window of ± 2 months around the
BIRD observation.
The SWIR signal was retrieved from the AATSR pixel where the GF was identified by
BIRD. For each observation, the signal is considered as being above the detection limit if its
value is above three times the background standard deviation. The background is a window of
approximately 10 × 10 km2 around the GF location pixel. The reported SWIR radiance value
is the radiance signal minus the background average.
The SWIR radiance signal thus retrieved shows a large variation throughout the time
window considered. Such a variation evidences temporal variation in the flaring
intensity, possibly GFs are either inactive or active, with varying levels of intensity,
throughout time. In a subsequent step, it is also determined whether the radiance
at the GF location pixel is a local maximum (after excluding other possible GF
locations within the surroundings). Then, a new average and standard deviation
for each GF may be constructed, more representative of when the GF is actually
on.
Those figures are then compared to a normalized RP. The normalization is necessary
because the BIRD fire processor reports parameters for clusters of cells which may
comprehend more than one reported GF within its borders.
The main conclusions are that the relationship between SWIR radiance and RP is
temperature dependent and that the relationship is approximately linear within six
temperature bins. |
|
|
|
|
|
|