Wavelet transforms allow for the decomposition of precipitation fields over a broad range of spatial scales. The wavelet coefficients obtained from the decomposition of daily global precipitation fields can be used to study the spatial characteristics of the model simulated field over multiple spatial scales, in addition to diagnosing the effective model resolution. At a given spatial scale, the variability of the corresponding wavelet coefficients is proportional to the amount of “details” needed to pass from the representation of the precipitation field at that scale to the representation with respect to the next -finer- scale. The greater the variability, the greater the details required, and the greater the energy associated with that specific spatial scale. If we consider a sequence of daily precipitation fields over several years, we can construct the energy spectrum of global precipitation and analyse its evolution with climate change. We also propose utilising wavelets in future downscaling activity.
We have applied a 2D Haar wavelet transform to the ERA5 global daily precipitation fields for the period from 1950 to 2020. Then, we have studied the variability of the daily wavelet coefficients for the different spatial scales. If we compare the two normal periods, 1961-1990 and 1991-2020, it can be seen that for the most recent period there is a shift of the maximum of the energy spectrum towards smaller scales. The shift is more pronounced in the tropics. If we consider the time series of the energies, there is an increase in the energies for most of the spatial scales after 1985. The growth rates among the scales are different, though, and precipitation at the Meso-beta scale and the lower part of the synoptic scale (up to around 440 km) has become more important in the total wavelet energy balance of daily precipitation. The wavelet analysis enables us to detect a change in the scale structure of the global daily precipitation patterns with climate change.
The results presented here are part of the work described in detail in the scientific article:
How to cite: Lussana, C., Casati, B., Benestad, R. E., Lutz, J., Dobler, A., Landgren, O., Haugen, J. E., Mezghani, A., and Parding, K. M.: Wavelet transform applied to ERA5 global daily precipitation fields to assess changes in the rainfall patterns, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-181, https://doi.org/10.5194/ems2022-181, 2022.
