본문컨텐츠영역
AMI Cal/Val System
The ACVS (AMI Cal/Val System) of the National Meteorological Satellite Center is a comprehensive quality management system for GK2A. Referring to the NOAA ICVS in 2021, a pilot system was built and development was completed in 2022. Since then, web services have been provided starting from 2023 after the pilot operation. The GK2A Status displays dynamic graphs of the payload status information, radiometric quality information, and geometric quality information. The existing GSICS inter-calibration page has been expanded and revamped to provide services for Infra-red channels, Visible channels including not only Earth targets but also GSICS Methods, Ray-matching, and Lunar calibration information. In the future, we will perform sensor information for subsequent satellites, add new reference satellite data, and develop calibration technologies to provide integrated quality information along with the meteorological satellite data operated by the NMSC.
GK2A radiometric and geometric calibration quality
The quality indicators for the radiometric calibration meteorological data of the GK2A satellite in the visible channel (VNIR) are expressed in terms of Signal-to-Noise Ratio (SNR), while for the infrared channels (MWIR to LWIR), they are expressed in terms of Noise Equivalent Differential Temperature (NEdT). The threshold values for these indicators are shown in Table 2.4.
Table 1. The radiometric calibration quality threshold values for the GK2A satellite| Band | SNR | Band | NEdT | Band | NEdT | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| @240K | @300K | @240K | @300K | |||||||
| VNIR | VIS04 | 261 | MWIR | SW038 | 2.70K | 0.18K | LWIR | IR096 | 2.70K | 0.18K |
| VIS05 | 299 | WV063 | 0.40K | 0.10K | IR105 | 0.40K | 0.10K | |||
| VIS06 | 130 | WV069 | 0.37K | 0.10K | IR112 | 0.37K | 0.10K | |||
| VIS06 | 300 | WV073 | 0.32K | 0.10K | IR123 | 0.32K | 0.10K | |||
| NIR13 | 300 | WV087 | 0.27K | 0.10K | IR133 | 0.27K | 0.10K | |||
| NIR16 | 300 | |||||||||
Visible channel: Signal-to-Noise Ratio(SNR)
The signal-to-noise ratio (SNR) of the GK2A AMI visible channel detector can be calculated using the following equation (2.1), which is the ratio of the Solar Calibration Target (SCT) radiance at 100% albedo to the total noise.
Infrared channels: Noise Equivalent Differential Temperature (NEdT)
The Noise Equivalent Differential Temperature (NEdT), which is a radiometric performance metric for the infrared channels, is obtained by converting the range of radiance errors that correspond to noise to a temperature range, according to the Planck function and conversion conditions. The temperature variation within the error range corresponding to noise can be obtained for both low and high temperatures, using a space-based reference temperature of 240K for very cold temperatures and a black body reference temperature of 300K for high temperatures. Equation 2.2 is the expression for the Noise Equivalent Differential Temperature. It is the ratio of the radiance equivalent to the noise that is induced by ICT observations to a fixed coefficient in a temperature-radiance conversion table.
Positioning Accuracy
The image position correction performance of the GK2A satellite can be evaluated for each visible channel using the star residual value. The quality metric uses the 3σ value of the residual for each visible channel observed around the Earth, and the reference values are shown in Table 2.9.
Table 2. Image location correction quality standard value| classification | Accuracy requirement(μrad) | |
|---|---|---|
| East-west direction | North-south direction | |
| Positioning Accuracy | 42 | 42 |
Visible Calibration
The validation of the visible and near-infrared channels of the GK2A Advanced meteorological imager (AMI) is performed using an absolute calibration device called the Solar Diffuser. However, if any anomalies occur in the satellite or the Solar Diffuser, normal calibration cannot be performed. Therefore, continuous calibration using methods such as inter-comparison with low-orbit satellites or lunar calibration should be performed in conjunction with the absolute calibration device.
The ray-matching method is a technique for directly comparing the values obtained from high-precision low-earth orbit and geostationary satellites that have observed the same region, in order to achieve high accuracy in calibration and validation. When comparing the values, the solar and observation angles are limited to reduce differences in the characteristics of the two satellites, and errors due to the observation angle of the geostationary satellite and differences in spectral response functions (SRF) are corrected to perform more accurate comparisons.
Moon inter-calibration: GK2A satellite performs approximately 40 moon observations each month for radiometric performance analysis and monitoring using its visible near-infrared channel. Moon observations are conducted when the Moon is near the Earth, utilizing the LA area observation of GK2A. The observed lunar radiance is compared with the lunar black model (GIRO) to analyze changes in the observed values.
Infrared Calibration
GK2A satellite is designed to maintain a certain level of accuracy, but during operation, the performance of the sensors may vary due to optical adjustments or changes in sensitivity. Therefore, it is necessary to continuously monitor and improve the quality of the sensors. To do so, international collaborative research is being conducted using the Global Space-based Inter-Calibration System (GSICS) to mutually test and monitor the quality of the satellite.
inter-calibration between low Earth orbit satellites (GEO-LEO)
The inter-calibration of infrared channels is performed using reference satellite (MetOp-B/-C IASI, SNPP/NOAA20 CrIS) data from well-characterized low Earth orbiting (LEO) satellites' hyperspectral infrared sounders. It is carried out on 10 channels with central wavelengths ranging from 3.8μm to 13.3μm through spatio-temporal, response function matching, and spatial homogeneity tests. The daily biases are calculated from the standard brightness temperatures using regression analysis for 29 days (target date ±14 days) of data.