Dear Zahra Mokatari,

If it is research purpose, you may study the effect of green infrastructure on urban cooling/heating, but if your purpose is to making policy advices, you may have to understand that the urban green infrastructure possibly in few instances, has detrimental effect on the air pollution.

Although there is no clear understanding of the pollution effects due to insufficient data, there is chance, especially in the place you are from shows.

Cheers,

Kishore

Yes you are right these researches are always along with high uncertainty, thank you very much for taking time

Dear Zahra,

this review provides an overview of evapotranspiration measurement methods: Article Measurement and estimation of actual evapotranspiration in t...

... and also this contribution from the University of Arizona:

http://www.swhydro.arizona.edu/archive/V7_N1/feature3.pdf

Best,

Sascha

you can use land surface temperature data from satellite to see the variation of green space.

https://www.sciencedirect.com/science/article/pii/S1618866717301541

To get an evapotranspiration extent of your city, you can do urban climate simulation for your city. This will give you not only data on latent heat flux, sensible heat flux and heat storage but also helps you to study the causality, boundary layer depth etc.

Dear Gemechu,

I think the modeling fluxes are useless until you validate with the observed fluxes. Without having them validated, the urban climate (e.g., causality, boundary layer depth, UHI, etc ) studies are not that useful.

Cheers,

Kishore

It is enough to validate some variables and the others are derived from these variables. You can't validate all vars. The derivations are based on physics and they are as realistic as the validated ones. I don't know what you mean by "not useful" unless you don't have good knowledge of the basic sciences.

Dear Gemechu,

I got my PhD on vegetation fluxes! I know about the level of physics is in the land surface models. Many processes are overly parameterized. I understand that all variables can't be validated but you must know that the surface fluxes are starting point of any atmospheric simulation, not other way around, and also remember that all of the surface fluxes are synthesized using surface pressure, precipitation, air temperature, air humidity, winds, radiation input, and unfortunately all these are strongly controlled by the surface. Therefore, you should not skip to validate these parameters while studying atmospheric phenomenon.

Hope it is now clear why I am saying the validation is needed.

Regards,

Kishore

As far as I know most models are validated and their performance is good in many respects for the current climate. Th e level of accuracy is dependent on many factors, such as the quality of input fields (initail and boundary) and complexity of the land surface features. That doesn't mean that the models are not useful. Ofcourse they are much more useful than even measurements/observations. Sometimes even models predict much better than observations. If projection is required, that is another issue. Weather and Climate are chaotic and it is quite difficult to accurately say when and where extreme events occur. For policy purposes, model validation may not be enough, intercomparisons of the mo dels is also important. For more information refer to this paper Article Review of urban surface parameterizations for numerical climate models

"As far as I know most models are validated and their performance is good in many respects for the current climate." Good. I suggest you to compare surface fluxes obtained from various land surface models. You can get this data from the NASA, and all these models use same input data, and analyse those fluxes, then you will realise what I am saying.

"Th e level of accuracy is dependent on many factors, such as the quality of input fields (initail and boundary) and complexity of the land surface features." I agree with it, but even if the input and surface features are represented correctly, the over-parameterizations mislead the flux calculations.

"That doesn't mean that the models are not useful. " I never said this. There are inherent issues in the models so one needs to validate the models with observations.

"Ofcourse they are much more useful than even measurements/observations. Sometimes even models predict much better than observations.". This does not make any sense. Check again if I am wrong !!!

"For policy purposes, model validation may not be enough, intercomparisons of the mo dels is also important." I totally agree with you. Policymaking is according to the present level of knowledge, you can safely go for the model inter-comparison, but the research is for going forward. As a researcher, I think about the present capabilities and future improvements.

Finally, I will check the reference paper, but I don't accept every article foolishly; I read critically.