Jack Xiong

An overview of the Earth Observing System MODIS instrument and associated data systems performance

The MODIS instrument on the EOS Terra Mission has completed over 2 years of successful operation. Excellent data products have been developed and a full year or more of these products are now available. Validation of these products is continuing and efforts to improve product availability and access are underway. The MODIS on the EOS Aqua satellite is projected to become operational in 2002.

Achieving satellite instrument calibration for climate change

For the most part, satellite observations of climate are not presently sufficiently accurate to establish a climate record that is indisputable and hence capable of determining whether and at what rate the climate is changing. Furthermore, they are insufficient for establishing a baseline for testing long-term trend predictions of climate models. Satellite observations do provide a clear picture of the relatively large signals associated with interannual climate variations such as El Nino-Southern Oscillation (ENSO), and they have also been used to diagnose gross inadequacies of climate models, such as their cloud generation schemes. However, satellite contributions to measuring long-term change have been limited and, at times, controversial, as in the case of differing atmospheric temperature trends derived from the U.S. National Oceanic and Atmospheric Administrations (NOAA) microwave radiometers.

Establishing the Antarctic Dome C community reference standard site towards consistent measurements from Earth observation satellites

Establishing satellite measurement consistency by using common desert sites has become increasingly more important not only for climate change detection but also for quantitative retrievals of geophysical variables in satellite applications. Using the Antarctic Dome C site (75°06′S, 123°21′E, elevation 3.2 km) for satellite radiometric calibration and validation (Cal/Val) is of great interest owing to its unique location and characteristics. The site surface is covered with uniformly distributed permanent snow, and the atmospheric effect is small and relatively constant. In this study, the long-term stability and spectral characteristics of this site are evaluated using well-calibrated satellite instruments such as the Moderate Resolution Imaging Spectroradiometer (MODIS) and Sea-viewing Wide Field-of-view Sensor (SeaWiFS). Preliminary results show that despite a few limitations, the site in general is stable in the long term, the bidirectional reflectance distribution function (BRDF) model works well, and the site is most suitable for the Cal/Val of reflective solar bands in the 0.4–1.0 µm range. It was found that for the past decade, the reflectivity change of the site is within 1.35% at 0.64 µm, and interannual variability is within 2%. The site is able to resolve calibration biases between instruments at a level of ∼1%. The usefulness of the site is demonstrated by comparing observations from seven satellite instruments involving four space agencies, including OrbView-2–SeaWiFS, Terra–Aqua MODIS, Earth Observing 1 (EO-1) – Hyperion, Meteorological Operational satellite programme (MetOp) – Advanced Very High Resolution Radiometer (AVHRR), Envisat Medium Resolution Imaging Spectrometer (MERIS) – dvanced Along-Track Scanning Radiometer (AATSR), and Landsat 7 Enhanced Thematic Mapper Plus (ETM+). Dome C is a promising candidate site for climate quality calibration of satellite radiometers towards more consistent satellite measurements, as part of the framework for climate change detection and data quality assurance for the Global Earth Observation System of Systems (GEOSS).

Correction of subframe striping in high resolution MODIS ocean color products

The MODIS (Moderate Resolution Imaging Spectroradiometer) scanner makes subframe measurements in some of its bands to increase the spatial resolution from its standard 1km resolution to 500m or 250m. This is achieved by sampling a detector of a high resolution band at twice (or four times) the sampling rate of the 1km bands. This paper shows that a calibration equation nonlinear with radiance and specific to the individual subframes will reduce striping in the images. The effects are significant for low radiance levels like those encountered over ocean scenes. A preliminary calibration correction is derived with two approaches: first from prelaunch measurements, then from on-orbit data. The results of the two methods are qualitatively similar.

MODIS pre-launch reflective solar band response vs. scan angle

The MODIS scan mirror reflectance is a function of angle of incidence (AOI). For the MODIS solar reflective bands (RSB), it is specified that the calibrated response variation versus scan angle (RVS) should be less than 2% and the uncertainty of the RVS characterization should be less than 0.5% within the scan angle range of -45° ~ +45°. During MODIS pre-launch RVS calibration and characterization, a series of laboratory tests were performed to assess the relative response versus scans angle for all MODIS bands. Utilizing a Spherical Integrating Source, SIS, as an illumination source, the test data was collected at various angles of incidence. The characterization of the RVS included a measurement uncertainty assessment, repeatability analysis, RVS modeling and determination. The results show good repeatability on the order of less than 0.5% for all the near infrared (NIR) bands and the visible (VIS) bands. The detector response variation across scan angles for the majority of the NIR and VIS bands meets the instrument specification. The derived RVS model enabled appropriate implementation of on orbit calibration. This paper summarizes the methodologies and the algorithms used in the MODIS pre-launch RVS calibration for the RSB bands, illustrates detector response variation with scan mirror angle of incidence, and demonstrates instrument specification compliance within the scan angle coverage of ±55 degree. As a result, the RVS model and the correction coefficients developed in the pre-launch calibration have been adopted during the MODIS on-orbit calibration.

Using CEOS reference standard test sites to track the calibration stability of NOAA-19 AVHRR reflective solar channels

In recent years, there is an increasing interest to establish a global integrated network of calibration sites for the purpose of tracking sensor performance, conducting cross-sensor comparison and assessing data quality and consistency. Based on such a need, the Committee on Earth Observation Satellites (CEOS) proposed eight instrumented sites for which surface measurements can be acquired through field campaigns and five pseudo-invariant desert sites typically consisting of sand dunes. In this study, we select one site from each category to study the calibration stability of reflective solar channels of NOAA- 19 Advanced Very High Resolution Radiometer (AVHRR) (launched on February 6, 2009). Since AVHRR does not have an onboard calibrator for the reflective solar channels and vicarious calibration often needs long-term observations to derive reliable trends, this study will provide an early assessment of sensor on-orbit calibration performance and establish a preliminary trend to examine its calibration consistency with other sensors. The Antarctic Dome C site is selected primarily to monitor the on-orbit calibration performance whereas Libya 4 test site is used to evaluate the cross-calibration consistency of AVHRR with other sensors. A site-specific Bi-directional Reflectance Distribution Function (BRDF) model developed based on observations made by Moderate Resolution Imaging Spectroradiometer (MODIS) is used to normalize AVHRR observed Top-of-Atmosphere (TOA) reflectances. Impact due to calibration applied to NOAA-19 AVHRR L1B is assessed separately using a constant detector response. Results show that for NOAA-19 AVHRR solar channels 1 and 2, variations in reflectance during the first year after launch are still around 6% and more than 10%, respectively, either due to sensor change or improper adjustment of calibration coefficients. While two sites provide consistent trends for the visible channel, the Dome C site is more suitable for the near-infrared channel as impacts of the absorption by atmospheric water vapor are minimal.

Calibration of low gain radiance at VIIRS emissive band (M13) and VIIRS image about moon temperature

Early assessment of the VIIRS thermal emissive bands (TEBs) show that all VIIRS TEBs except M13 are stable and exceed the specifications. M13 is a dual gain band, and is used for determining the surface temperature at low radiance (high gain) and fire detection at high radiance (low gain). At a low gain stage, the onboard blackbody temperature at an operational temperature of 292 Kelvin is far below the lowest temperature at the low gain, which prevents from any attempt to radiometric calibration. This study found that the VIIRS calibration data during the blackbody temperature warm up and cool down (WUCD) may be useful to check the gain stability and to estimate noise equivalent deviation of temperatures (NEdT). During the VIIRS blackbody temperature warm up and cool down, the blackbody temperature was cooled down to 267 K and warmed up to 315 K. The contrast at the low gain for M13 band between blackbody and space views may be useful, although the highest blackbody temperature is still below the low boundary for the low gain. Moon surface temperature can be as hot as 400 Kelvin, high enough for M13 band radiometric calibration at a low gain. The advantages using the observation data of Moon are that it is very stable and there is no gaseous absorption. However, Moon surface emissivity for infrared spectrum needs to be known. This study found that Moon may be used to check measurement range of the VIIRS M13 band at a low gain. We have developed a calibration algorithm to determine the moon temperature and generated the first VIIRS image about moon temperature. There are some other sources such gas flares that may also be used to estimate the radiometric accuracy at low gain.

A preliminary study of Aqua/MODIS snow coverage continuity with simulated band 6

Snow cover is one of the sensitive indicators of global climate change. Numerous studies have shown the importance of accurate measurements of snow cover. The Moderate Resolution Imaging Spectroradiometer (MODIS) is well suited to the measurement of snow cover because snow characteristically has high reflectance in the MODIS Visible (VIS) and low reflectance in the MODIS Shortwave Infrared (SWIR) wavelengths, a characteristic that allows for snow detection by a normalized ratio of VIS and SWIR bands. The automated MODIS snow-mapping algorithm uses at-satellite reflectance in MODIS VIS band 4 (0.545-0.565 μm) and SWIR band 6 (1.628-1.652 μm) to calculate the Normalized Difference Snow Index (NDSI). Aqua MODIS band 7 (2.105-2.155 μm) instead of band 6 has been used to calculate NDSI, in response to band 6 striping problem caused by non-functional or noisy detectors. In our early study, a feasible algorithm to map Aqua MODIS band 6 based on the relationship between Terra MODIS bands 6 and 7 has been developed and validated. This algorithm has been used to retrieve Aqua MODIS band 6. Aqua MODIS NDSI values computed from Aqua MODIS observed band 6, simulated band 6, and observed band 7 are used to map snow based on current MODIS snow algorithm, respectively. Snow coverage mapped using NDSI computed from observed band 6 is regarded as a standard snow product, comparison and analysis are performed between snow mapping using NDSI computed from simulated band 6 and observed band 7. This paper will investigate the measurement continuity between Terra and Aqua MODIS snow coverage products, and propose another alternative for Aqua MODIS NDSI retrieval. Our approach for monitoring snow coverage is valuable to keep the continuity and consistency for MODIS snow products.

Integrating the cross-sensor calibration and validation system for GEOSS support

The Global Earth Observation System of Systems (GEOSS) will provide long-term data for a wide variety of communities. . To be meaningful and useful in the societal benefit applications, the global satellite observations need continuous and consistent measurements, which require improved cross instrument calibration and product validation. Cross-sensor calibration/validation is necessary for achieving data continuity and consistency. Several institutes and laboratories have been trying to plan and initiate the appropriate processes to accomplish the calibration/validation indicated for the various measurements and disciplines that will most likely be involved. EastFIRE Laboratory at George Mason University (GMU) in the United States has been working on a cross-sensor calibration/validation system during the past seven years, and has demonstrated the capability and performance of the system for NPP/NPOESS prelaunch testing support. The EastFIRE cross-sensor calibration and validation system can be further extended to include more sensors and measurements to support the GEOSS communities in data consistency control and construction of long-term consistent Climate Data Records (CDRs). The primary objectives of the present system expansion and integration are: 1) to support satellite observation research and operations; 2) to support multiple sensor cross-calibration and product cross-validation; 3) to support prelaunch testing and post-launch validation of the next generation earth observation missions; and 4) to build global FCDRs (Fundamental Climate Data Records) for the GEOSS communities using Sensor Data Records (SDRs) from multiple sensors.

Impact of MODIS SWIR band calibration improvements on Level-3 atmospheric products

The spectral reflectance measured by the MODIS reflective solar bands (RSB) is used for retrieving many atmospheric science products. The accuracy of these products depends on the accuracy of the calibration of the RSB. To this end, the RSB of the MODIS instruments are primarily calibrated on-orbit using regular solar diffuser (SD) observations. For λ <0.94 μm the SD’s on-orbit bi-directional reflectance factor (BRF) change is tracked using solar diffuser stability monitor (SDSM) observations. For λ <0.94 μm, the MODIS Characterization Support Team (MCST) developed, in MODIS Collection 6 (C6), a time-dependent correction using observations from pseudo-invariant earth-scene targets. This correction has been implemented in C6 for the Terra MODIS 1.24 μm band over the entire mission, and for the 1.38 μm band in the forward processing. As the instruments continue to operate beyond their design lifetime of six years, a similar correction is planned for other short-wave infrared (SWIR) bands as well. MODIS SWIR bands are used in deriving atmosphere products, including aerosol optical thickness, atmospheric total column water vapor, cloud fraction and cloud optical depth. The SD degradation correction in Terra bands 5 and 26 impact the spectral radiance and therefore the retrieval of these atmosphere products. Here, we describe the corrections to Bands 5 (1.24 μm) and 26 (1.38 μm), and produce three sets (B5, B26 correction = on/on, on/off, and off/off) of Terra-MODIS Level 1B (calibrated radiance product) data. By comparing products derived from these corrected and uncorrected Terra MODIS Level 1B (L1B) calibrations, dozens of L3 atmosphere products are surveyed for changes caused by the corrections, and representative results are presented. Aerosol and water vapor products show only small local changes, while some cloud products can change locally by >10%, which is a large change.