12:00 〜 12:15
[ACG38-12] EarthCARE/MSI 雲プロダクトにおけるSmile Effectの定量的評価
キーワード：衛星観測、雲特性リトリーバル、EarthCARE衛星プロジェクト、スマイルエフェクト、NICAM/J-Sim模擬データ
During recent works, we have found a situation that MSI sensitivity coefficient could largely shift (Smile Effect) among some specific wavelengths (band1 and band3).
To evaluate the impact of the smile effect on cloud retrieval products, we chose four detector pixels from band1 and band3 (as well as the nadir pixel for reference) in order to elucidate how much the variation of sensitivity coefficient affects the optical thickness and effective radius of cloud droplet, by using Comprehensive Analysis Program for Cloud Optical Measurement (CAPCOM).
For the Typical case of shallow water clouds (τ=8, re=8μm), the smile effect for the cloud retrieval is not significant for most cases (up to 6% of error).
For the Typical case of deep convective clouds (τ=8, re=40μm), the smile effect for the cloud retrieval is even less significant for most cases (up to 4% of error).
Meanwhile, in order to see how the smile effect affects actual satellite observation scenes, we select a typical shallow water cloud scene and a deep convective cloud scene from NICAM/J-Sim simulation data, then get the retrieval of cloud properties using CAPCOM. The result matched quite well with it in the forward radiation simulation, indicating that the error from the smile effect is within 10% for most cases.
To evaluate the impact of the smile effect on cloud retrieval products, we chose four detector pixels from band1 and band3 (as well as the nadir pixel for reference) in order to elucidate how much the variation of sensitivity coefficient affects the optical thickness and effective radius of cloud droplet, by using Comprehensive Analysis Program for Cloud Optical Measurement (CAPCOM).
For the Typical case of shallow water clouds (τ=8, re=8μm), the smile effect for the cloud retrieval is not significant for most cases (up to 6% of error).
For the Typical case of deep convective clouds (τ=8, re=40μm), the smile effect for the cloud retrieval is even less significant for most cases (up to 4% of error).
Meanwhile, in order to see how the smile effect affects actual satellite observation scenes, we select a typical shallow water cloud scene and a deep convective cloud scene from NICAM/J-Sim simulation data, then get the retrieval of cloud properties using CAPCOM. The result matched quite well with it in the forward radiation simulation, indicating that the error from the smile effect is within 10% for most cases.