The measurement of many environmental variables using the remote sensing are made through the correlation with the information derived from these space and airborne instruments. As stated Don Enrique Cardenas-Sanchez is possible the correlation of surface temperature with atmospheric pressure, though, if required high accuracy rates is necessary to determine other variables can help us obtain such accuracy when calculations are made with the help of the models used.
Yes.It is possible.It greatly depends on the temporal and spatial resolution you need.As far as I know, products from satellites such as GOSAT ,OCO-2 can retrieve surface pressure together with CO2 concentration .But it is hard to cover the globe at all time.Model simulation results such as ECMWF will be a good choice to cover all.
As far as I know there are no direct measurement of air or atmospheric pressure through remote satellite sensing. I have been trying to find such satellite service myself for one of my atmospheric photon monitoring projects, but so far i could not find one that do direct measurements. My search is still on as far as I am concerned.
On the possibility, it think it is highly possible ... but it has to be correlational. All air pressure measurements has to be immersed in the atmosphere to be able to measure it with a high degree of accuracy, hence the term air pressure. All others have to do it indirectly taking into consideration all other intermediary factors or variables. It is like if we can not make a direct connection between variables A and C, perhaps we can find variable B. If A = B, and B = C, then perhaps, theoretically A = C.
If A = satellite imagery, and C = atmospheric pressure, then perhaps B = some other factor that has direct measurable and quantifiable links to both A and C. It could be temperature, infrared or UV readings, laser or radiographic images, sonars, cloud patterns and others which are measurable by sensor readings from kilometers up beyond the earth's atmosphere where air pressure = 0. Remote sensing can perform accurate measurements by parameters which can go through physical barriers in the atmosphere, like photons or electromagnetic waves. For example, cloud patterns like that of a thypoon or tornado, can be observed by satellites.
Since it is correlational, there must be numerous independent samples, thousands perhaps, of any B parameter for it to have have any statistical significance, at .001, .01 or .05 level, on A or C. Such huge statistical crunching can best be done with R software, the one used basically with scientifically oriented statistics. SPSS may be used to certain degree. Hypotheses on air pressure and sensor readouts by satellite imagery, may be established and proven to a certain degree of confidence.
Perhaps satellite sensors can be linked to terrestial sensors, realtime ... but this will take lots of time and work, linking thousands of sensors, arduinos, MCUs, programming computers and servers ... on a regional or worldwide basis. This could be done, as to what entity will do it on that scale is a big question. But perhaps the issue of global warming can push nations and individuals to link up and ... save this planet from a global weather catastrophe.
Meanwhile, we have to do our work with simple home or academe brewed devices, programs and statistical interpolations .... to say a brown dot at this latitude and longitude that is changing colors from brown to red means something ... significantly.