Re: [arts-users] Using ARTS for atmospheric correction AMSR2 brigthness temperatures
Dear Robin, As Stefan wrote, we don't have much experience of the lowest frequencies you mentioned, but in https://www.atmos-meas-tech.net/12/6341/2019/ we used ARTS down to 7 GHz. We did not consider the atmospheric settings in detail, so I can not say anything about the absolute accuracy. Anyhow, based on what I know, to get you started I would say: * Among gases, only N2, O2 and H2O should matter. Other gases can be ignored. * For these frequencies you can avoid "line-by-line" and get full coverage by applying a continua model for N2, and full absorption models for O2 and H20. * I would recommend to include LWC (liquid water content), that at these frequencies can be considered as purely absorbing. * As the continua and absorption models are quite fast you can calculate absorption "on-the-fly" (instead of using a pre-calculated lookup-table, that is a more complex setup). * The basic settings to achieve the above are: # Agenda for scalar gas absorption calculation Copy(abs_xsec_agenda, abs_xsec_agenda__noCIA ) # on-the-fly absorption Copy( propmat_clearsky_agenda, propmat_clearsky_agenda__OnTheFly ) # Definition of species # abs_speciesSet( species = [ "N2-SelfContStandardType", "O2-PWR98", "H2O-PWR98", "liquidcloud-ELL07" ] ) # No transitions needed # abs_lines_per_speciesSetEmpty * Note that you must include N2, O2 and H2O vmrs, in vmr_field, in the order used in abs_species (values between 0 and 1). LWC shall be given in kg/m3. * Rain will have a significant impact, but is more complex to include in the simulations. In addition, ERA5 does not give you full information on rain, only rain of stratiform type is reported (convective rain is treated differently in the model and is not reported). And it could be questioned if the rain is placed correctly in place and time. Bye, Patrick On 2020-04-16 17:17, Stefan Buehler wrote: Dear Robin, the best is to write to arts_users for now. It will depend on the nature of the issues you run into who will get most involved on our side. The example from the classroom exercise calculates only the pure line-by-line gas absorption spectrum. Scientific issues here could be: - Should additional absorbing gases be included? - Which gas absorption continua should be added? (I am not so familiar with the lower end of your frequency range, so somebody should look into this a bit.) - Do you want to add absorption by cloud liquid water and/or scattering by rain? Perhaps negligible, again I have no good intuition for these low frequencies. But before looking into these refinements, let’s first see if you can get the pure line-by-line calculation running. Best wishes, Stefan On 16 Apr 2020, at 15:56, Robin van der Schalie wrote: Dear Stefan, Thank you for the quick response, much appreciated. I'll move forwards with installing the package and try to play around with the suggested exercise. To be sure of correctly applying ARTS for this specific problem, it would be great if someone can support us on this matter. I see a positive impact for both the temperature input used for the soil moisture retrievals (based on the 37 GHz) and the soil moisture retrievals from higher frequencies (e.g. SSMI 18GHz) within the climate data record. If this would be successfully implemented within the current algorithm (the Land Parameter Retrieval Model), I see no problem in involving you in the publication(s) that follow and I expect it will have a good outreach through the CCI SM project. Let me know if you recommend someone that I could contact or if you prefer to be involved yourself. Kind regards, Robin van der Schalie -- *dr. Robin van der Schalie* // Senior Remote Sensing Scientist VanderSat // Satellite observed water data. Globally. Daily. Wilhelminastraat 43a, 2011 VK, Haarlem, The Netherlands *T* +31 23 3690093 *M* +31 6 81631591 *W* www.vandersat.com -- Op do 16 apr. 2020 om 10:30 schreef Stefan Buehler < stefan.bueh...@uni-hamburg.de>: Dear Robin, yes, ARTS should be well suited for this. There even already is a python classroom exercise with a setup for computing and displaying optical depth. (On github, in package atmtools/arts-lectures, directory exercises/04-rtcalc .) What you may need some advice on is which absorption models to actually use (ARTS offers a lot of choices, I don’t remember if the ones in the exercise are the best for a real application). ARTS is free to use, the best reward for us is involvement in scientific publications. So, depending on how much support you will need, we would expect the person(s) that helped you to be included in the first publication on this. Best wishes, Stefan On 15 Apr 2020, at 13:05, Robin van der Schalie wrote: Good afternoon, My name is Robin van der Schalie and I am
Re: [arts-users] Using ARTS for atmospheric correction AMSR2 brigthness temperatures
Dear Stefan, Thank you for the quick response, much appreciated. I'll move forwards with installing the package and try to play around with the suggested exercise. To be sure of correctly applying ARTS for this specific problem, it would be great if someone can support us on this matter. I see a positive impact for both the temperature input used for the soil moisture retrievals (based on the 37 GHz) and the soil moisture retrievals from higher frequencies (e.g. SSMI 18GHz) within the climate data record. If this would be successfully implemented within the current algorithm (the Land Parameter Retrieval Model), I see no problem in involving you in the publication(s) that follow and I expect it will have a good outreach through the CCI SM project. Let me know if you recommend someone that I could contact or if you prefer to be involved yourself. Kind regards, Robin van der Schalie -- *dr. Robin van der Schalie* // Senior Remote Sensing Scientist VanderSat // Satellite observed water data. Globally. Daily. Wilhelminastraat 43a, 2011 VK, Haarlem, The Netherlands *T* +31 23 3690093 *M* +31 6 81631591 *W* www.vandersat.com -- Op do 16 apr. 2020 om 10:30 schreef Stefan Buehler < stefan.bueh...@uni-hamburg.de>: > Dear Robin, > > yes, ARTS should be well suited for this. There even already is a python > classroom exercise with a setup for computing and displaying optical > depth. (On github, in package atmtools/arts-lectures, directory > exercises/04-rtcalc .) What you may need some advice on is which > absorption models to actually use (ARTS offers a lot of choices, I > don’t remember if the ones in the exercise are the best for a real > application). > > ARTS is free to use, the best reward for us is involvement in scientific > publications. So, depending on how much support you will need, we would > expect the person(s) that helped you to be included in the first > publication on this. > > Best wishes, > > Stefan > > On 15 Apr 2020, at 13:05, Robin van der Schalie wrote: > > > Good afternoon, > > > > My name is Robin van der Schalie and I am currently the person in > > charge of > > running soil moisture retrievals based on passive microwave > > observations within the ESA Climate Change Initiative ( > > https://www.esa-soilmoisture-cci.org/). > > > > In our never ending search for ways to further improve our soil > > moisture > > retrieval algorithm, which is based on the Land Parameter Retrieval > > Model, > > I would like to get a better handle on the atmospheric effects that > > alter > > the AMSR2 (and other historical mission) brightness temperatures from > > ground level. In essence, having more realistic Atmospheric Optical > > Depth > > values. This would be for multiple frequencies, i.e. L-band (1.4 GHz), > > C-band (6.9 GHz), X-band (10.7 GHz), Ku-band (18 GHz), K-band (23 GHz) > > and > > Ka-band (37 GHz). For this I am already preparing a database from > > reanalysis (ERA5) on the water vapor, atmospheric pressure, and air > > temperature as input for the calculation. > > > > From going through the ARTS documentation it seems to me that this > > goal > > would be achievable using your package (especially the Typhon as we > > work > > with python). Is that a correct assumption? And if so, could you maybe > > provide me with some guidance on how to get started on this? > > > > Hope to hear from you soon, > > > > Robin van der Schalie > > > > > > > > > > > > -- > > *dr. Robin van der Schalie* // Senior Remote Sensing Scientist > > VanderSat // Satellite observed water data. Globally. Daily. > > Wilhelminastraat 43a, 2011 VK, Haarlem, The Netherlands > > *T* +31 23 3690093 *M* +31 6 81631591 *W* www.vandersat.com > > > > -- > > ___ > > arts_users.mi mailing list > > arts_users.mi@lists.uni-hamburg.de > > https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi > ___ arts_users.mi mailing list arts_users.mi@lists.uni-hamburg.de https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
Re: [arts-users] Using ARTS for atmospheric correction AMSR2 brigthness temperatures
Dear Robin, the best is to write to arts_users for now. It will depend on the nature of the issues you run into who will get most involved on our side. The example from the classroom exercise calculates only the pure line-by-line gas absorption spectrum. Scientific issues here could be: - Should additional absorbing gases be included? - Which gas absorption continua should be added? (I am not so familiar with the lower end of your frequency range, so somebody should look into this a bit.) - Do you want to add absorption by cloud liquid water and/or scattering by rain? Perhaps negligible, again I have no good intuition for these low frequencies. But before looking into these refinements, let’s first see if you can get the pure line-by-line calculation running. Best wishes, Stefan On 16 Apr 2020, at 15:56, Robin van der Schalie wrote: Dear Stefan, Thank you for the quick response, much appreciated. I'll move forwards with installing the package and try to play around with the suggested exercise. To be sure of correctly applying ARTS for this specific problem, it would be great if someone can support us on this matter. I see a positive impact for both the temperature input used for the soil moisture retrievals (based on the 37 GHz) and the soil moisture retrievals from higher frequencies (e.g. SSMI 18GHz) within the climate data record. If this would be successfully implemented within the current algorithm (the Land Parameter Retrieval Model), I see no problem in involving you in the publication(s) that follow and I expect it will have a good outreach through the CCI SM project. Let me know if you recommend someone that I could contact or if you prefer to be involved yourself. Kind regards, Robin van der Schalie -- *dr. Robin van der Schalie* // Senior Remote Sensing Scientist VanderSat // Satellite observed water data. Globally. Daily. Wilhelminastraat 43a, 2011 VK, Haarlem, The Netherlands *T* +31 23 3690093 *M* +31 6 81631591 *W* www.vandersat.com -- Op do 16 apr. 2020 om 10:30 schreef Stefan Buehler < stefan.bueh...@uni-hamburg.de>: Dear Robin, yes, ARTS should be well suited for this. There even already is a python classroom exercise with a setup for computing and displaying optical depth. (On github, in package atmtools/arts-lectures, directory exercises/04-rtcalc .) What you may need some advice on is which absorption models to actually use (ARTS offers a lot of choices, I don’t remember if the ones in the exercise are the best for a real application). ARTS is free to use, the best reward for us is involvement in scientific publications. So, depending on how much support you will need, we would expect the person(s) that helped you to be included in the first publication on this. Best wishes, Stefan On 15 Apr 2020, at 13:05, Robin van der Schalie wrote: Good afternoon, My name is Robin van der Schalie and I am currently the person in charge of running soil moisture retrievals based on passive microwave observations within the ESA Climate Change Initiative ( https://www.esa-soilmoisture-cci.org/). In our never ending search for ways to further improve our soil moisture retrieval algorithm, which is based on the Land Parameter Retrieval Model, I would like to get a better handle on the atmospheric effects that alter the AMSR2 (and other historical mission) brightness temperatures from ground level. In essence, having more realistic Atmospheric Optical Depth values. This would be for multiple frequencies, i.e. L-band (1.4 GHz), C-band (6.9 GHz), X-band (10.7 GHz), Ku-band (18 GHz), K-band (23 GHz) and Ka-band (37 GHz). For this I am already preparing a database from reanalysis (ERA5) on the water vapor, atmospheric pressure, and air temperature as input for the calculation. From going through the ARTS documentation it seems to me that this goal would be achievable using your package (especially the Typhon as we work with python). Is that a correct assumption? And if so, could you maybe provide me with some guidance on how to get started on this? Hope to hear from you soon, Robin van der Schalie -- *dr. Robin van der Schalie* // Senior Remote Sensing Scientist VanderSat // Satellite observed water data. Globally. Daily. Wilhelminastraat 43a, 2011 VK, Haarlem, The Netherlands *T* +31 23 3690093 *M* +31 6 81631591 *W* www.vandersat.com -- ___ arts_users.mi mailing list arts_users.mi@lists.uni-hamburg.de https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi ___ arts_users.mi mailing list
Re: [arts-users] Using ARTS for atmospheric correction AMSR2 brigthness temperatures
Dear Robin, yes, ARTS should be well suited for this. There even already is a python classroom exercise with a setup for computing and displaying optical depth. (On github, in package atmtools/arts-lectures, directory exercises/04-rtcalc .) What you may need some advice on is which absorption models to actually use (ARTS offers a lot of choices, I don’t remember if the ones in the exercise are the best for a real application). ARTS is free to use, the best reward for us is involvement in scientific publications. So, depending on how much support you will need, we would expect the person(s) that helped you to be included in the first publication on this. Best wishes, Stefan On 15 Apr 2020, at 13:05, Robin van der Schalie wrote: Good afternoon, My name is Robin van der Schalie and I am currently the person in charge of running soil moisture retrievals based on passive microwave observations within the ESA Climate Change Initiative ( https://www.esa-soilmoisture-cci.org/). In our never ending search for ways to further improve our soil moisture retrieval algorithm, which is based on the Land Parameter Retrieval Model, I would like to get a better handle on the atmospheric effects that alter the AMSR2 (and other historical mission) brightness temperatures from ground level. In essence, having more realistic Atmospheric Optical Depth values. This would be for multiple frequencies, i.e. L-band (1.4 GHz), C-band (6.9 GHz), X-band (10.7 GHz), Ku-band (18 GHz), K-band (23 GHz) and Ka-band (37 GHz). For this I am already preparing a database from reanalysis (ERA5) on the water vapor, atmospheric pressure, and air temperature as input for the calculation. From going through the ARTS documentation it seems to me that this goal would be achievable using your package (especially the Typhon as we work with python). Is that a correct assumption? And if so, could you maybe provide me with some guidance on how to get started on this? Hope to hear from you soon, Robin van der Schalie -- *dr. Robin van der Schalie* // Senior Remote Sensing Scientist VanderSat // Satellite observed water data. Globally. Daily. Wilhelminastraat 43a, 2011 VK, Haarlem, The Netherlands *T* +31 23 3690093 *M* +31 6 81631591 *W* www.vandersat.com -- ___ arts_users.mi mailing list arts_users.mi@lists.uni-hamburg.de https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi ___ arts_users.mi mailing list arts_users.mi@lists.uni-hamburg.de https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
