Spectral Characteristics of Convective-Scale Precipitation Observations and Forecasts

Abstract

As an alternative to traditional precipitation analysis and forecast verification, 1D and 2D spectral decompositions of NCEP/Stage IV and Multi-Radar Multi-Sensor (MRMS) precipitation products and convective-scale model forecasts are examined. Both the stage IV and MRMS analyses and the model forecasts show a similar weak power-law behavior in 1D spectral decompositions, although the MRMS analysis does not drop off in power at wavelengths less than approximately 20 km as found in the stage IV analysis. The convective-scale forecasts produce similar behavior to the MRMS when the forecast model’s effective resolution is sufficient. Neither the MRMS analyses nor the forecasts suggest the existence of a break in the spectral slope at the scales for which the analyses and forecasts are valid. The 2D spectra of both observations and forecasts, expressed in terms of an absolute wavenumber and azimuthal angle, show power varying significantly as a function of azimuthal angle for a given wavenumber. This azimuthal anisotropy is significant, and is dominated by the second mode (wavenumber 2). The phase of the mode is the result of the orientation of precipitation features and, hence, convective system orientations and propagation. Observations show a shift in orientation (phase) over May–June–July. The convective forecasts reproduce this shift in phase, although with a consistent but small phase error.

Using wavelets to verify the scale structure of precipitation forecasts

“…In the first step, we have studied the climatological properties of the local wavelet spectra. Similar analyses of the predicted average spatial structure were carried out by Willeit et al (2015) and Wong and Skamarock (2016) using Fourier transforms. Aggregation of these mean spectral properties according to the weather situation has confirmed the findings of Brune et al (2018), who report that wavelet spectra are very well suited to differentiate between rain fields with different degrees of spatial organization.…”

“…Overall this type of climatological analysis has proven to be a useful first evaluation of the average model performance. The natural possibility to localize errors in space constitutes an advantage over the Fourier approach of Willeit et al (2015) and Wong and Skamarock (2016).…”

A wavelet‐based analysis of convective organization in ICON large‐eddy simulations

“…A similar index to WOI may be obtained using two‐dimensional Fourier spectra (e.g. Wong and Skamarock, ). One major difference between Fourier and wavelet transforms is the localization of the latter.…”

“…Because direct three-dimensional observations of (thermo-)dynamical variables like wind, temperature or moisture are generally not available, convective organization is often characterized on the basis of radar or satellite measurements using Hovmöller diagrams (Carbone et al, 2002) or Fourier spectra (Wong and Skamarock, 2016). Other studies categorize convection using organization indexes like the simple convective aggregation index (SCAI; Tobin et al, 2012) or the index of organization (I org ; Tompkins and Semie, 2017).…”

Mesoscale Resolution Radar Data Assimilation Experiments with the Harmonie Model

“…Matching the radar resolution to that of the model should minimize numerical effects while focusing on physical features. In particular, the belt and cell features in precipitation fields [22,23] are better simulated with RDA. Higher vertical resolution used in "cube-smoothing" RDA allows us to correct the position of the condensation level, which reduces the phase error.…”