No abstract

Earth's land-sea distribution modifies the temperature response to orbitally induced perturbations of the seasonal insolation. We examine this modification in the frequency domain by generating 800,000-yr time series of maximum summer temperature in selected regions with a linear, two-dimensional, seasonal energy balance climate model. Previous studies have demonstrated that this model has a sensitivity comparable to general circulation models for the seasonal temperature response to orbital forcing on land. Although the observed response in the geologic record is sometimes significantly different than modeled here (differences attributable to model limitations and feedbacks involving the ocean-atmosphere-cryosphere system), there are several results of significance: (1) in mid-latitude land areas the orbital signal is translated linearly into a large (>10°C) seasonal temperature response; (2) although the modeled seasonal response to orbital forcing on Antarctica is 6°C, the annual mean temperature effect (<2°C) is only about one-fifth that inferred from the Vostok ice core, and primarily restricted to periods near 41,000 yr; (3) equatorial regions have the richest spectrum of temperature response, with a 3000-yr phase shift in the precession response, plus some power near periods of 10,000–12,000 yr, 41,000 yr, 100,000 yr, and 400,000 yr. Peaks at 10,000–12,000 yr and 100,000 and 400,000 yr result from the twice-yearly passage of the sun across the equator. The complex model response in equatorial regions has some resemblance to geologic time series from this region. The amplification of model response over equatorial land masses at the 100,000-yr period may explain some of the observed large variance in this band in geologic records, especially in pre-Pleistocene records from times of little or no global ice volume.

A one-level seasonal energy balance climate model which has explicit two-dimensional land-sea geography is introduced and its properties analyzed. The model distinguishes land from sea surface strictly by the local thermal inertia employed. The transport is governed by a smooth latitude dependent diffusion mechanism. The model's seasonal cycle is solved for, with and without ice feedback. In each case, good agreement with real data is accomplished with minimal tuning of the adjustable parameters. The model sensitivity for small solar constant changes is in agreement with that of other accepted models. However, when the solar constant is lower a few percent, the ice cap grows discontinuously from covering Greenland and vicinity to a size three times larger and mostly into Northern Canada. A similar discontinuous growth occurs when the orbital elements are changed to favor cool summers in the Northern Hemisphere. This discontinuous sensitivity is discussed in the context of the Milankovitch theory of the ice ages, and the associated branch structure is shown to be analogous to the so called 'small ice cap' instability known from simpler models. Although the discontinuity in sensitivity is somewhat conjectural, both linear and nonlinear models have their largest sensitivity to orbital parameter changes in the northern portions of the large continents; in the western hemisphere this also corresponds to the ice sheet boundary, suggesting that the land-sea configuration is crucial to large glacier growth irrespective of moisture budgets. Other experiments and implications are also discussed.

No abstract

Abstract. We present the zonal mean temperature variations for the quasi-biennial oscillation (QBO) and the semiannual oscillation (SAO) based on data from SABER on the TIMED spacecraft (years 2002 to 2004) and from MLS on the UARS mission (1992 to 1994). The SABER measurements provide the rare opportunity to analyze data from one instrument over a wide altitude range (15 to 95 km), while MLS data were taken in the 16 to 55 km altitude range a decade earlier. The results are presented for latitudes from 48 • S to 48 • N. New results are obtained for the QBO, especially in the upper stratosphere and mesosphere, and at mid-latitudes. At Equatorial latitudes, the QBO amplitudes show local peaks, albeit small, that occur at different altitudes. From about 20 to 40 km, and within about 15 • of the Equator, the amplitudes can approach 3.5 • K for the stratospheric QBO (SQBO). For the mesospheric QBO (MQBO), we find peaks near 70 km, with temperature amplitudes reaching 3.5 • K, and near 85 km, the amplitudes approach 2.5 • K. Morphologically, the amplitude and phase variations derived from the SABER and MLS measurements are in qualitative agreement. As a function of latitude, the QBO amplitudes tend to peak at the Equator but then increase again pole-ward of about 15 • to 20 • . The phase progression with altitude varies more gradually at the Equator than at mid-latitudes. Many of the SAO results presented are also new, in part because measurements were not previously available or were more limited in nature. At lower altitudes near 45 km, within about 15 • of the Equator, the temperature amplitudes for the stratospheric SAO (SSAO) reveal a local maximum of about 5 • K. At higher altitudes close to the Equator, our results show separate peaks of about 7 • K near 75 and 90 km for the mesospheric SAO (MSAO). In the SAO results, significant inter-annual differences are evident, with the amplitudes be- for the QBO, the SAO temperature amplitudes go through minima away from the Equator, and then increase towards mid latitudes, especially at altitudes above 55 km. We compare our findings with previously published empirical results, and with corresponding results from the numerical spectral model (NSM). Although not a focus of this study, we also show results for the inter-annual variations (which appear to be generated at least in part by the QBO) of the migrating diurnal tide. In the upper mesosphere, their amplitudes can approach 20 • K, and they are derived jointly with the zonalmean components.