Relationships between meteorological data and soil properties: quantifying precipitation and aridity in the Middle Danube Basin through geophysical proxies

Spatial patterns of precipitation and aridity across Europe are most likely to vary in response to changing temperature. Our knowledge about such responses is, however, still limited as most geoscientific studies provide point data rather than cover wider areas. Such an approach would require reliable proxies, which can be determined in high spatial resolution along with detailed knowledge about how these reflect climate. Classically, we derive (paleo-) precipitation and -temperature for terrestrial areas from climofunctions, which base on magnetic susceptibility (χ) and its frequency dependence (χ_fd). These parameters reflect the quantity and modification state of magnetic minerals like magnetite, which are dependent on the combined influence of temperature and precipitation. Recently, also the maghemite contribution to the high-temperature dependent susceptibility (χ_td) has been used to create climofunctions, which are mainly constrained to reconstruction of (palaeo-) precipitation. Yet, such relationships have mostly been reported from Asia.

In this study, we test if we can qualify and quantify (palaeo-) precipitation, temperature and aridity by room- and high-temperature rock magnetic and colorimetric data of recent topsoils in a narrow precipitation range between ~535 mm/a and 585 mm/a. The data are derived from geographically evenly distributed bulk-samples from the Backa Loess Plateau (Middle Danube Basin). Our results show that we can quantify precipitation by rock magnetic properties (χ, χ_fd, as well as χ_td), but colorimetric methods are more challenging to interpret. While we can also reconstruct aridity, temperature is difficult to determine in a meaningful way.

In this contribution, we show the results of statistical analysis performed on a multivariate rock magnetic and colorimetric dataset, and their relation to geographical differences in prevailing climatic regimes of the Middle Danube Basin. While care needs to be taken not to overfit the data (due to more colorimetric variables than spatial data points), we regard our multivariate approach at least as relevant as trying to fit individual magnetic or colorimetric proxies to climate variables.