All celebrative pies have been cooked and eaten. The PhD-allowance-budget-friendly champagne, popped. The very first published paper of my PhD – and its various media declinations – overly Tweeted, LinkedIned, ResearchGated. What was missing was this blogpost.

In our paper, our main message is that it is the local context of an area that influences the tools, the drought indicators, chosen to qualify droughts and their severity. But it is not the same thing as measuring the severity of a drought. Those indicators only take into consideration hydro-climatic variables and they overlook the human processes like, for example, the vulnerability to drought of the exposed area.

But what do we mean by hydro-climatic variables? Applied to drought, they are the physical variables measuring the deviations from the expected average of the water content in the components of the water cycle. The values of these physical indicators are attributed to three categories of droughts: meteorological, agricultural and hydrological. A meteorological drought indicator will measure a deficit of rainfall. Agricultural drought indicators will measure a deficit of water in the soils and vegetation. Hydrological drought indicators will measure a deficit in the river flows and the lake or reservoir levels. It is commonly what we think about when drought comes to our mind: less rain, scorched soils and dry rivers. What can also come to our mind when thinking about droughts are famines, water supply shortages, water-related conflicts, migrations, loss of livelihood. And the list goes on. Those are common drought-related impacts. These impacts testify to the gravity of drought on the ground. But generally, impacts and their associated indices are not included in drought monitoring.

In a way, it is understandable that such drought impacts are not monitored or associated to indicators. The main reason is that there is no consensus on how to define an impact. Drought impacts are non-structural, they can be direct or indirect depending on how long a drought persists in time or the size of the area it affects. All these uncertainties make them difficult to quantify, measure and monitor.

However, basing drought monitoring on merely physical variables frames at best only half of the situation. In the same way, formulating drought-related interventions on the basis of this monitoring can lead to incomplete responses. In that sense, in our study, we wanted to question and understand the linkage and separation between drought drivers and drought impacts. To this purpose, we reviewed more than 5000 scientific studies reporting on drought drivers and drought impacts for affected countries and analysed how these two compared.

This is what we observed: Global North countries focus on reporting drivers of drought, whereas drought-related food and water insecurities are more reported in Sub-Saharan Africa and Australia-Oceania, respectively. Beyond this first striking observation, four reasons explain why countries report one feature of drought more than the others:

First is the physical availability of water. Drought driver indices measure the water deficit in one or several of the components of the hydrological cycle. For countries with an arid climate, rainfall and surface water resources are already scare. It thus makes more sense in these climates to measure a water deficit at the crop level and in that way, use indicators related to agricultural drought.

Second,  the socio-economic conditions of the countries will determine the means available to invest in drought research and monitoring networks. Food and water securities can be affected by drought but not only. They can also be affected by rapid population growth, governance, political instability and corruption for instance. In other words, focusing on physical drivers of drought is an advantage more apt to be of interest in areas where more basic and essential needs, such as food security, have been met.

Third, data availability affects the selection and accuracy of an index, especially if the chosen index is unsuitable for the particular climate. Indeed, not all countries benefit from a robust hydro-climatic historical database. Generally, rainfall is the most monitored variable with however, no guarantee of a robust and continuous historical dataset that is needed for an accurate calculation of a rainfall-based index. Misuses include its calculation even with unadvised gaps in the dataset or inappropriate applications to arid climates qualified by inherent long periods without rain. Fourth is the scientific approach and the interest in the country that determines from which physical and/or social science scope drought will be looked at and for what purpose. Most of the drought driver indices are investigated under a more physically based approach. Food and water securities related to drought, respectively more reported in Sub-Saharan Africa and in Australia–Oceania, are also studied through a physical perspective but they are also approached through the lens of social sciences. In many western countries, institutional incentives may favour research that falls into well-defined silos. Research that meaningfully incorporates both physical and social sciences may not be sufficiently interesting to merit ground-breaking publications on both fronts; it may instead require one or the other discipline serving in a more consultative role. On the other hand, food and water security are holistic concepts that also involve socio-economic processes as mentioned earlier, but also human processes such as conflicts, equality and water justice to only cite a few.

To sum up, it is the use and selection of an index that will determine the incidence of a type of meteorological, agricultural or hydrological drought rather than its effective incidence. As rainfall is the variable that is easier to rely on, monitor and compute into a drought index, drought is mainly approached merely as a rainfall statistical anomaly and equated to meteorological drought. Focusing on the drivers of drought is nonetheless an advantage that northern, and generally wealthier countries, can afford, as they have met essential food and water needs. More importantly, the scientific reporting of drought drivers is not linearly connected to the reporting of drought-related impacts, nor are the drought-related food and water securities studies linearly connected. This suggests that drought impacts reporting is certainly dependent on both the local physical and human processes occurring in the geographic area.

This advocates for more research in which the scope of drought mitigation is widened to the vulnerability to drought events rather than only their probability of occurrence. Monitoring the severity of drought more accurately could be reached by also including drought indices that are people-centred. These indices could measure to what extent populations are exposed and vulnerable to having their welfare hampered by drought. Hopefully, this further research is what I aim to do during the rest of my PhD. Optimistically, this will also be leading to celebrative pies, sparkling beverages and ostentatious dissemination of the results on the World Wide Web.