Hi Patrick， Thanks for the quick reply! Regarding the first point mentioned, the simulation is described in detail as follows. The first simulation example based on is the artscomponents/montecarlo.arts file, with the input of 3-dimensional ERA5 contour data including temperature, humidity, pressure, and altitude, and the absorption model including water vapor, nitrogen, and oxygen, with the water vapor and oxygen being MPM93 models. Then according to the product matching with the Microwave Thermometer of Fengyun3 Satellite (MWTS) to the height and line of sight of the sensor in flight, input to sensr_pos and sensor_los, the sensor setup adopts sensor_responseGenericAMSU. other agendas include the surface agendas as well as the non-refractive situation of the air. The surface is modeled using TELSEM for land and TESSEM for ocean depending on the situation. However, the simulation results are highly biased, in the teens of K or more. Can you please see where the problem is? Or is there any related work that you can recommend as a reference?
One more question, I would like to ask if the OEM method of ARTS2.4 is perfect, I didn't use QPACK of matlab because I am more familiar with ARTS language. Looking forward to your reply. Best wishes. Sincerely, ZhangChao > -----原始邮件----- > 发件人: "Patrick Eriksson" <[email protected]> > 发送时间: 2023-09-26 22:15:47 (星期二) > 收件人: "张超" <[email protected]>, [email protected] > 抄送: > 主题: Re: [arts-users] consultancy question > > Hi, > > Thanks for your interest in ARTS. > > > Unfortunately, I have observed a significant deviation between the > > simulated bright temperatures and the actual measurements, with > > discrepancies reaching up to 20 K. I suspect that there may be an issue > > with my approach. Have you encountered similar research or do you have > > any insights into what the problem might be? > > This is impossible to answer without further information. Please provide > more details. > > > > > Secondly,the HITRAN database has a clearly defined purpose for its use, > > whereas Perrin, which is frequently utilized, has a mixed source,could > > you have a reference for this? > > For AMSU and to get started, it is easier to use some of the predined > absorption models for O2, N2 and H2O, by e.g. Rosenkranz. > > > > Thirdly,when representing the Jacobi matrix for gases in a graph, I > > often express it in "rel" units, but I am uncertain about the > > appropriate notation. For example, would it be preferable to represent > > it as K (kg/kg)? > > It is for a relative change. So maybe just best to avoid the second > part. An option is to divide with 100 and write K/%. > > Bye, > > Patrick </[email protected]></[email protected]>
