Power spectra and coherence of drought in the interior plains

Abstract

Mean growing season (April-September) precipitation series at 407 stations in the Interior Plains of North America, dating back to 1985 in some stations, are used to derive a Bhalme and Mooley-type drought index for the region. Linear trend, representing first order non-stationarity, is first removed from the drought series. Then the periodogram via the Fast Fourier Transform (FFT), Maximum Entropy (ME) and Non-Integer (NI) spectral techniques are applied to search for significant periodicities in the temporal characteristics of the drought series. Cross-spectral technique is also used to characterize the extent of spectral coherence in the drought series.It is found that no significant power in drought is present at periods greater than about 6 years. Spectra of drought in the Interior Plains are dominated by three peaks in the quasi-biennial, quasi-triennial and quasi-5-year oscillations. The periodic 18.6-year lunar nodal cycle obtained by others for the Drought Area Index (DAI) for the region is completely absent in the spectra of the analysed drought index series. Also the well-known 11-year sunspot cycle is only evident in the drought index series for a few of the stations. In general there is an ill-defined drought recurrence interval.Moreover the areal extent of peaks contained within these oscillations do not indicate any significant spatial coherence. These periodicities can be evident at one station while being absent at another a few hundred kilometres away. On the basis of the spectral results, it is concluded that obtained significant periodicities are only short-lived time variabilities in drought as recorded by individual stations and do not represent any persistent drought characteristics over a broad geographical region. A cross-spectral analysis of the drought series with sunspot numbers shows no evidence to indicate a significant relation between drought and the single sunspot cycle. In general, the growing season drought series in the Interior Plains displays only what appears to be a random variation, with the possible superimposition of the quasi-biennial oscillation.KEY WORDS Drought Periodogram maximum entropy and non-integer spectral techniques Drought spectral coherence Cycles

Characterizing the spatial and temporal variability of June–July moisture conditions in the Canadian prairies

“…Cook et al (1997) also found evidence of a bidecadal drought rhythm in western North America since 1700. Despite the long-standing debate about whether solar or lunar cycles are responsible for these oscillations (Oladipo, 1987;Padmanabhan and Ramachandra Rao, 1990;Cook et al, 1997;Cerveny and Shaffer, 2001), no conclusive evidence exists (Thejll, 2001). No bidecadal oscillation was found in the June-July moisture spectra, and the 10-15 year peak is too broad to be linked directly to the 11 year sunspot cycle.…”

“…Quasi-biennial oscillations (QBOs) are a common feature in many time series from around the world, such as: precipitation (Oladipo, 1987), tree rings (Briffa et al, 1990), air temperature (Mason, 1976), SST (Allan, 2000), NAO (Hurrell, 1995;Allan, 2000), snow cover (Kane, 2000), and OLR (Kane, 2000). The cause of these QBOs is often attributed to reversals in stratospheric winds.…”

“…Alternatively, the QBO peak in the June-July moisture spectra on the Canadian prairies may be caused by oscillations in eastern equatorial Pacific SSTs, since Kane (2000) found evidence of both quasi-biennial (2-3 years) and quasi-triennial (3-4 years) oscillations. Periodicities of 3-8 years have commonly been associated with El Niño-southern oscillation (ENSO) variability (Oladipo, 1987;Allan, 2000;Kane, 2000). ENSO periodicities are somewhat ill defined because they have changed over time.…”

Luni‐solar 18.6‐ and 10—11‐year solar cycle signals in H. H. lamb's dust veil index

“…It should be noted that only 50 spectral estimates are given in Figure 2(a and b), but 500 could have been computed in which case the spectra would be continuous curves. Following Currie's (1989) critique of Oladipo's (1987) improper procedures on North American rainfall records, CWO present numerous examples showing that fitting linear or higher order polynomials to data, and using the residuals to compute a spectrum, usually results in failure to detect the M, and S, terms. Figure 2(c) presents the spectrum of the residuals for dviA after a linear polynomial was fitted to the data by least squares.…”

On studying relations between time series in climatology

“…The goal is to describe time series in the time and frequency domains, including climatic teleconnections that can hardly be found within the correlation/regression approach. Although methods of multivariate time series analysis are also described, mostly in the frequency domain, in climatology (von Storch and Zviers, 1999) and oceanography (Emery and Thomson, 2004), they are rarely applied in both time and frequency domains (e.g., Privalsky, 1988;Privalsky and Jensen, 1995) or just in the frequency domain (e.g., Oladipo, 1987;Schneider and Schönwiese, 1989;Ghanbari et al, 2009;Park and Dusek, 2013).…”