Interactions between the meteorology and hydrology communities – Interview with Martin Best

Contributed by Maria-Helena Ramos, member of the Irstea Guest Columnist Team

At Irstea, the Catchment Hydrology group working in Antony (southern Paris) has been developing research that aims at better integrating knowledge from meteorology in hydrological models and forecasting systems.

In December last year, I met Martin Best, an expert scientist on land surface modelling at the MetOffice, during a workshop organized in Lillehammer by the Norwegian Water Resources and Energy Directorate (NVE), the Norwegian Meteorological Institute (MET), and the University of Oslo (UiO).

Martin is the leader of a group working on land surface processes at both the MetOffice in Exeter and at the CEH in Wallingford (Joint Centre for Hydro-Meteorological Research, JCHMR). Particularly, he leads the development of the Joint UK Land Environment Simulator (JULES), which is a community land-surface model.

We had the opportunity to briefly discuss about the need for more interactions between the meteorology and hydrology communities in modelling and forecasting. Martin kindly accepted to share some of his thoughts with the Hepex community, as reported here below.

Maria-Helena Ramos: When talking about the variety of models currently available for applications, you have stated that “if you have to have different models for different model applications, your modelling approach is probably wrong.” Could you tell us more about it? Can it be an obstacle to coupling land-surface atmospheric models with hydrological models?

Martin Best: If we think about the real world, then there is no distinction in physics between, say, mean flows and peak flows in hydrology. The same physical processes apply all the time. So if you need to use different models for differing applications, then I think we could say that the models do not have an ideal representation of the real world. This is not something that is unique to hydrological models though. In the land surface modelling community we have said for a long time now that we need to make it clear to people that you can not take soil moisture from one land surface model and just put it in another. The models have different meanings for their soil moisture. However, there is only one definition of soil moisture in reality, so these models are not complete in their representation of the physical processes.

So why do we have this model diversity? Well, in defence of our models I would say that the reason is because it is not possible to observe the soil moisture that we are trying to model. Observations of soil moisture come from point measurements, but these do not represent the heterogeneity of the land surface or horizontal scales that we are modelling, or from satellite data, but this only gives us the soil moisture in the top few centimetres of the soil. Hence we can not evaluate our models properly. In addition, land surface models have traditionally been designed to represent evaporation from the surface as a boundary condition to atmospheric models, whereas hydrological models have been designed to represent stream flows. None of the models have closed the terrestrial water cycle by being evaluated over all three components (evaporation, stream flow and changes in the water stores, e.g., soil moisture and groundwater).

I think it is only by evaluating our models (both land surface and hydrological) over all components of the terrestrial water cycle that we will identify the limitations of our modelling approaches and start to address them. I believe that if we did this we would identify the strengths and weaknesses of the two modelling communities and demonstrate that we can learn a lot from each other.

MHR: It seems that collaboration in hydrometeorological modelling between the two communities of meteorology and hydrology is still a challenge. How do you think we could facilitate this?

MB: At present I think we look on ourselves as two separate communities tackling two different problems, but in truth, we are all trying to model the same thing: the terrestrial water cycle. Historically there may have been a scale gap in our focus between global land surface modelling and hydrological catchment modelling that could have been a contributing factor to a lack of integration. This is now changing and new collaborations are emerging. However, we still have differing approaches to our modelling and often use different language. So there are some remaining challenges that need to be overcome to enable effective collaboration.

In terms of what we could do to facilitate working together in the future, I think there are a few things we need to consider. Firstly, we need to have a better understanding of the motivations and requirements within each community. We need to recognize that we all have strengths and weaknesses, but to use a common phrase, by coming together we can be more than the sum of the parts.

However, I think the biggest challenge that we face is a psychological one. I feel that at the moment we are all focused on our own issues and challenges and as such, what information we need to address these. However, to make real advances we need to look on ourselves as just one part of the larger jigsaw. Hence we need to change from “this is what I need” to “this is what I can offer”. It sounds like a small change, but it will be difficult for all of us.

MHR: Any final comments or messages to the hydrological forecasting community?

MB: I think we face a number of challenges for the future. We are living in a climate that does not seem to be stationary, so for instance, does it make sense to talk about return periods when the long term statistics that are needed to calculate these have trends? Yet resilience to flooding is as big a challenge as it has ever been. So society needs us to develop a complete end to end modelling system that we can have confidence in, to give us accurate information on the global and, perhaps more importantly, terrestrial water cycle.

So let’s see if we can all work together with a common purpose of increasing our understanding, and improving our predictive capabilities for what happens to the water that is on the land surface, and the subsequent impacts that this has on all of us.

Thank you, Martin, for your time and contribution!

• More about Martin Best and his work can be found here.
• And if you want to have a look at Martin’s presentation, “Organization of and research focus in hydrometeorological modelling in UK“, you can download it from here.