17:15 〜 18:45
[ACG36-P12] 水蒸気・気温推定を目的とした地上テラヘルツ放射計のチャネル最適化と数値実験による逆解析結果
キーワード：リモートセンシング、ラジオメータ、逆推定、テラヘルツ観測
The frequency of radio waves that extends from 100 GHz to 10 THz is called the terahertz (THz) region. Currently, there is still a band gap between microwave and infrared observations, and THz radiometer (THR) observations are expected to bring advantages to atmospheric observations, such as estimation of thin cloud properties and soundings of water vapor profiles. It will be a help to get a better understanding of cloud which is regarded as one of the least understood components of the weather and climate (Kuzi, 2001). Several pioneering studies exist using a THR from aircraft and satellites that have successfully obtained atmospheric properties with unique sensitivity to the THz region (Fox, 2017; Wu, 2019). While the primary scientific motivation for a THR development is to retrieve ice cloud properties from above the sky, we are specifically focused on a ground-based radiometer to improve accuracy of near-surface temperature and water vapor mixing ratio (WVMR) profiles by the intense absorption effects on water vapor in the THz band. It has been noted that monitoring equivalent potential temperature (EPT) from the surface to 1 km is important to understand and predict the life cycle of severe weather events (Kato, 2011). EPT is related to atmospheric temperature and WVMR, and a more accurate assessment of these parameters in the lower atmosphere would contribute to atmospheric science and weather forecasting. For this purpose, we consider simultaneous observation using a ground-based microwave radiometer (Radiometrics Model MP-3000) and a THR with multi-channels around 300 GHz. It is given that the THR is designed to sample radio wave at least two channels in 310 to 350 GHz and the noise equivalent delta temperature (NEDT), which indicates the accuracy of the observations, will be less than 0.1 K using the latest THz-band RF technology.
In designing THR, it is important to select the best channel for our motivation. In this study, we first selected two channels of THR, which were evaluated to be the most effective in terms of information content of WVMR at lower troposphere. Among several optimization methods, we applied the Interactive Method (IM), which is a method based on information theory that evaluates the reduction of background uncertainty by making observations and can select the most effective observation channels for estimating temperature and WVMR profiles (Rogers, 1998). We used the U.S. Standard Atmosphere (USSA) as our profile and selected the best two channels in the range of 310 GHz to 350 GHz under clear-sky (CS) conditions with no water droplets or ice clouds. In this approach, the diagonal component of the background error covariance matrix (BECM) for temperature was set to 5% of the USSA profile and the diagonal component of the WVMR to 50%. The channels selected under these conditions were 341.5 GHz and 310.5 GHz [Figure 1]. It should be noted that changing assumptions (atmospheric profile, BEMC, NEDT, etc.) will lead to different results for optimized channel selection. Secondly, a numerical simulation was performed to obtain temperature and WVMR using the MP-3000 and an optimized THR at the same time. In this simulation, the line by line radiative transfer model (LBLRTM) was used as the forward model, and the maximum a posteriori (MAP) solution was applied to obtain the temperature and WVMR profiles. In this case, the atmospheric profile is assumed to be under CS conditions. The retrieved results suggest that the simultaneous measurement of MP3000 and the optimized THR improve the accuracy of WVMR and air temperature estimation, especially in the lower atmosphere, compared to the MP3000 alone [Figure 2].
In this study, we selected two optimal THz channels under certain assumed conditions and obtained results suggesting that simultaneous observation of MP3000 and THR improves the accuracy of near-surface WVMR and temperature. In the future, we plan to apply the channel selection and inversion methods developed in this study to actual THR observations to verify their evaluation.
In designing THR, it is important to select the best channel for our motivation. In this study, we first selected two channels of THR, which were evaluated to be the most effective in terms of information content of WVMR at lower troposphere. Among several optimization methods, we applied the Interactive Method (IM), which is a method based on information theory that evaluates the reduction of background uncertainty by making observations and can select the most effective observation channels for estimating temperature and WVMR profiles (Rogers, 1998). We used the U.S. Standard Atmosphere (USSA) as our profile and selected the best two channels in the range of 310 GHz to 350 GHz under clear-sky (CS) conditions with no water droplets or ice clouds. In this approach, the diagonal component of the background error covariance matrix (BECM) for temperature was set to 5% of the USSA profile and the diagonal component of the WVMR to 50%. The channels selected under these conditions were 341.5 GHz and 310.5 GHz [Figure 1]. It should be noted that changing assumptions (atmospheric profile, BEMC, NEDT, etc.) will lead to different results for optimized channel selection. Secondly, a numerical simulation was performed to obtain temperature and WVMR using the MP-3000 and an optimized THR at the same time. In this simulation, the line by line radiative transfer model (LBLRTM) was used as the forward model, and the maximum a posteriori (MAP) solution was applied to obtain the temperature and WVMR profiles. In this case, the atmospheric profile is assumed to be under CS conditions. The retrieved results suggest that the simultaneous measurement of MP3000 and the optimized THR improve the accuracy of WVMR and air temperature estimation, especially in the lower atmosphere, compared to the MP3000 alone [Figure 2].
In this study, we selected two optimal THz channels under certain assumed conditions and obtained results suggesting that simultaneous observation of MP3000 and THR improves the accuracy of near-surface WVMR and temperature. In the future, we plan to apply the channel selection and inversion methods developed in this study to actual THR observations to verify their evaluation.