Environmental Engineering Reference
In-Depth Information
more reliably by applying similar algorithms for high-level data products, and it is
also possible to obtain error bounds of data among multiple sensors, which are
critical in climate change detection and analysis.
Sponsored by Northrop Grumman Space Technology (NGST), EastFIRE Labora-
tory at George Mason University has been working on cross-sensor comparison/
validation/calibration for NPP/NPOESS support since 2003. The hyper-spectral
measurements of the Atmospheric Infrared Sounder (AIRS) onboard NASA satellite
Aqua were used to simulate thermal emissive bands of MODIS and VIIRS. Band
mapping algorithms, as well as software toolkits and testing database, have been
developed for MODIS, AIRS comparison, and VIIRS proxy data generation study
(Hao et al.
2005a
,
b
,
2007
;Quetal.
2005a
,
b
,
2006a
,
b
; Hao and Qu
2009a
). The
capability and performance of these approaches have been validated by comparing
AIRS-simulated global MODIS SDR and aggregated MODIS SDR at AIRS footprints
(Hao et al.
2005a
,
b
). The band mapping approaches can also be used for generating
global FCDRs (Hao and Qu
2008
,
2009b
). In this chapter, we focus on the thermal
emissive SDRs at the 11- and 12-
m channels, which are used for surface temperature
retrieval. Technical approaches are described in details, and quantitative relationships
between thermal emissive measurements of different sensors are investigated and
discussed toward the construction of thermal emissive FCDRs.
μ
5.2 Data and Technical Methods
The AIRS (
http://airs.jpl.nasa.gov
) onboard NASA Aqua satellite is a hyper-
spectral sensor in the thermal infrared region. With 2,378 spectral channels,
AIRS has high spectral resolution and can provide more accurate information of
the atmosphere. AIRS measurements can be used as a bridge to evaluate the spectral
differences of various broadband sensors for different missions, such as AVHRR,
MODIS, and VIIRS. By convolving the spectral response functions of broadband
sensors with AIRS hyper-spectral measurements, proxy data for AVHRR, MODIS,
and VIIRS will be generated. Then, a global database can be generated including
proxy datasets and AIRS scene characteristics such as satellite zenith angle, satel-
lite azimuth angle, surface type, cloud fraction, etc. Based on analysis of the global
testing database, statistical relationship between different sensors, i.e., band transfer
equations, can be derived to convert consistent measurements for these sensors.
Details of band mapping approach for FCDR generation are illustrated in Fig.
5.3
.
The main steps include:
1. Collection of AVHRR/MODIS/VIIRS sensor specification and global AIRS
measurements
• Collect the spectral response functions of AVHRR band 4 and band 5,
MODIS band 31 and band 32, and VIIRS band M15 and band M16.
• Collect global AIRS L1B measurements for selected 8 days during
2002-2008 and in different seasons: 09/06/2002, 01/25/2003, 01/26/2003,
Search WWH ::
Custom Search