Most of my past research has attempted to understand the physics controlling the Earth's climate.
I have done quite a lot of work on the hydrological cycle. Early work was on detection and attribution. It's no longer a central interest, but it does inform a lot of the statistical work that we do. For example, Joe Osborne has looked at rainfall change in the supposedly best observed Northern Hemisphere mid-latitude region and found that observations apparently do not produce a reduction in rainfall in response to aerosol forcing. This is something of a surprise given that it is a robust response of modelled simulations (Osborne and Lambert, 2014). Joe's current work is looking at what is going on in runoff observations.
Figure 1: Cartoon distribution of land masses in the tropics by relative humidity and surface temperature rank. Locations in the top right corner are rainy; others largely aren't. Thanks to the Met Office Design Studio for the pic.
Changes in regional precipitation in the tropics under modelled climate change are dominated by shifts in patterns of rainfall rather than intensification or weakening of pre-existing patterns. Lambert et al., (2017) and Todd et al., (2018) developed a simple mean field theory scheme that shows how shifts in rainfall are associated with the relative ranks of temperature and relative humidity of geographical locations. The scheme is reasonably successful, but also gives us a baseline where it doesn't work for investigating the effects of atmospheric circulation.
Meanwhile, as far as we know, global mean precipitation is controlled by the conservation of energy, which imposes a stricter constraint on small changes in precipitation than does moisture availability. In Lambert and Webb, (2008), we argued that the reason that % / K changes in precipitation are more robust than the value of climate sensitivity across models is because shortwave cloud feedbacks affect the tropospheric energy budget relatively little. Hence, overall, the 1 -- 3 % / K changes in precipitation produced by models seems unlikely to match the biggest changes reported by some satellite observations. A remaining uncertainty is the effect of longwave cloud feedbacks, which can affect the tropospheric energy budget, and therefore precipitation. In current models, however, it doesn't seem that it would be possible to produce longwave cloud feedbacks that produce very different changes in precipitation and probably not without making large differences to climate sensitivity of temperature (Lambert et al., 2013). We await the new work on convective aggregation!
Under quite globally uniform forcing, surface temperature responses behave like local responses to the global mean radiative forcing, even away from equilibrium. In models, this means that simulated patterns of temperature change look similar under different forcing scenarios. In the real world, it means that land and ocean temperatures continue to increase in a constant ratio that we do not expect to change if warming accelerates or if it slows down. This is true even though the land shows a much smaller heat capacity to climate change than the ocean.
The constant ratio and similar patterns occur because the atmosphere is very good at transporting heat, particularly around a latitude circle. This makes detecting the influence of poorly understood regional aerosol forcings difficult, because their influence is not strongly regionally confined. Regional forcings do have slightly different temperature signatures, however, even large-scale sulphate aerosol forcing, as Manoj Joshi has convinced me.
This means that one can detect regional forcings (old news, I know), although surface temperature response might not be the best way to go about it. In fact, it might be the atmospheric energy transport that we should be trying to trace, perhaps through its impact on the hydrological cycle. This is part of the reason Joe Osborne looked at aerosol impacts on European rainfall.
The flipside of all this, is that solar radiation management geoengineering schemes that rely on regional forcing or forcing that doesn't have the same spatial signature as CO2 forcing (all as far I'm aware) will work. Because they will stimulate changes in atmospheric energy transport, however, they might stimulate nasty or at least unpredictable changes in the hydrological cycle.
I've done a fair bit of research on radiative forcing and feedbacks, especially short-term adjustment to forcing, but mostly not for a long time. Nevertheless, at present, I'm answering a challenge from Patrick Taylor at NASA Langley to explain the pattern of the regional tropical water vapour and lapse rate feedbacks. These don't show a sharp anti-correlation like the tropical mean water vapour and lapse rate feedbacks across climate models, but they are largely explained by simple processes. It's nearly done.
Finally, a personal perspective on my favourite paper ever. Lambert and Faull, (2007) described the basic difference in global mean response between carbon dioxide and solar forced climate model experiments: tropospheric adjustments deal with forcing absorbed by the atmosphere by reducing precipitation in the carbon dioxide case and increasing outgoing longwave in the solar case. It took four years to get it published, the results were mostly written-up by others before I got it out there and the results that weren't published by others were better reproduced elsewhere afterwards. (Also the altitude hypothesis for tropospheric adjustment in the final paragraph is almost certainly wrong.) But... it was the first piece of work I ever did on my own. Having the idea, testing it and realising that it was right convinced me to become a professional scientist. Whether this is good for me or anyone else is unclear. But if you find yourself in a similar situation, then I hope it offers some encouragement.