Faulty Climate Models: “Drought Forecasts Are Barely Trustworthy”

  • Date: 21/04/16
  • Axel Bojanowski, Spiegel Online

Global warming is supposed to produce more droughts. But now an analysis reveals that climate models can barely calculate rainfall.

Scientists predict a fundamental change in the climate, because humankind constantly emits greenhouse gases into the air. The gases retain heat in the air so that the temperature rises.

The greatest risk involves the water cycle between soil, air and sea: Global warming would boost this cycle as more water would evaporate. Water vapor is also a greenhouse gas; if its levels increase in the air, it would enhance the warming.

What’s more, an accelerated water cycle, paradoxically, would intensify droughts – as shown in calculations of climate models on which climate change forecasts are based.

A new study, however, questions these very models. Climatologists lead by Fredrik Charpentier Ljungqvist of Stockholm University have analysed historical climate data. In the case of rainfall, the data contradict the results of climate models, the researchers report in the science journal Nature.

For the period of man-made climate change in particular the computer models incorrectly stimulate the real changes in precipitation.

The study poses a great challenge for climate research, according to Matthew Kirby of California State University in a Nature commentary on the new paper. Eduardo Zorita, co-author of the study, explains in the interview, what the findings mean for the credibility of climate predictions.

Interview with Eduardo Zorita

Eduardo Zorita studies the climate of past millennia. He is a renowned expert on climate models and historical climate data. Dr Zorita works at the Helmholtz Centre for Materials and Coastal Research in Geesthacht near Hamburg.

Eduardo Zorita

SPIEGEL ONLINE: Mr. Zorita, you have disturbing news for your colleagues: The climate models that predict how the climate will change in the future seem to have significant shortcomings according to your research, right?

Eduardo Zorita: Well, the forecast that the air and the oceans will continue to warm up in response to man-made greenhouse gases is not in question. With regards temperatures the models seem to be working well. But our study shows that the climate models have problems with calculating changes in precipitation.

SPIEGEL ONLINE: That would go to the heart of climate predictions because the most important predictions relate to changes in precipitation. What should we think of the warnings of more drought events?

Zorita: These forecasts are untrustworthy. Our work shows that the results of the climate models differ evidently from the climate data for precipitation.

SPIEGEL ONLINE: How did you find out?

Zorita: We analysed changes in moisture conditions in the entire northern hemisphere during the last 1,200 years, based on the growth of tree rings, deep sea sediments or stalactites.

SPIEGEL ONLINE: Every year growing trees add a new ring, its trunk is slightly wider. In rainy years, the rings are thicker, in drought years they are thinner. But in addition to the effect of rain, there are also other environmental influences that effect the annual tree rings – why do you still trust their data?

Zorita: We have checked for plausibility by evaluating the climate data independently. It turned out that they provide individually similar results as in the overall analysis. In addition, the interpretation of climate data is derived from other studies that were made by colleagues who work independently from us and we have now united the data into a survey.

SPIEGEL ONLINE: Do the data confirm that the climate has already changed the way it has been calculated by the models? Precipitation too would have to show a change during the 20th century due to human influences.

Zorita: Our data do not show any abnormalities during the 20th century; precipitation nothing really changed. From the ninth to the eleventh century it was similarly dry, but then there was no man-made climate change. Even severe droughts like the recent one in the western United States are tempered by similar data from the Middle Ages. In addition, the data of the last 1200 years show that rainfall appears to be more volatile than was previously thought.

SPIEGEL ONLINE: Dry regions are predicted to become drier according to the climate models, damp regions are predicted to become wetter – these are the warnings of climate predictions. Can you at least confirm this assumption?

Zorita: No, we can’t. Although the scenario is physically plausible, we do not see it in the data.

SPIEGEL ONLINE: It is considered basic knowledge that warming temperatures boost the water cycle, i.e. more water will evaporate in response. Water vapour as a greenhouse gas would thus lead to an acceleration of global warming. Is that also wrong?

Zorita: No, the assumption seems plausible for the future, provided that the warming intensifies. However, for the last 1200 years we don’t have detected any evidence for a link between global warming and changes in rainfall. This is food for thought.


Zorita: Our study is a warning signal. It shows that we need to test climate models better. They can hardly model the water cycle which is the central climatic phenomenon.

SPIEGEL ONLINE: Are you not afraid that you will now draw the concentrated criticism of climate modelers?

Zorita: We hope there will be criticism – that’s what keeps science alive. Our study is not comprehensive after all; for example, it is only based on information about the northern hemisphere – and even there are some areas that are only lightly covered. And we cannot make any statements about extreme rainfall since our data show only moisture averages over several years. Nevertheless, we consider our results to be an urgent commission to fill gaps in our knowledge.

SPIEGEL ONLINE: What kind of gaps are you thinking of?

Zorita: We must better explore the behavior of clouds and airborne particles, the so-called aerosols. Even our understanding of how moisture is exchanged between ground and air is insufficient. This is problematic because it is these factors that largely determine the future of the climate.

Full interview (in German)

Translation GWPF