Person:

Rhines, Andrew Nelson

Errors in the timing assigned to observations degrade estimates of the power spectrum in a complicated and nonlocal fashion. It is clear that timing errors will smear concentrations of spectral energy across a wide band of frequencies, leading to uncertainties in the analysis of spectral peaks. Less understood is the influence of timing errors upon the background continuum. We find that power law distributions of spectral energy are largely insensitive to errors in timing at frequencies much smaller than the Nyquist frequency, though timing errors do increase the uncertainty associated with estimates of power law scaling exponents. These results are illustrated analytically and through Monte Carlo simulation and are applied in the context of evaluating the power law behavior of oxygen isotopes obtained from Greenland ice cores. Age errors in layer counted ice cores are modeled as a discrete and monotonic random walk that includes the possibility of biases toward under- or overcounting. The δ(^{18}O_{ice}) record from the Greenland Ice Sheet Project 2 is found to follow a power law of (1.40 \pm 0.19) for periods between 0.7 and 50 ky, and equivalent results are also obtained for other Greenland ice cores.

Severe impacts result from extreme events such as heat waves, droughts, cold spells, and floods. Characterizing and predicting variations in climate that give rise to these phenomena is important for mitigating their effects on human and natural systems. This thesis investigates whether climate variability is measurably changing and describes the observational basis for recent shifts in the temperature distribution. New methodology is presented that robustly estimates local distributional changes and permits for mapping them to regional or global scales, overcoming limitations of previous analyses. Contrary to the widespread view that climate variability has increased in recent decades, these analyses show that temperature variability has generally declined — albeit with important regional differences.

Historical observations of temperature are crucial for long-term monitoring of Earth's climate. However, hundreds of millions of daily observations contain precision-related biases that prevent their use in distributional analyses. A new machine-learning algorithm automatically corrects for these biases, enabling their use in long-term climate studies. The algorithm increases the number of usable observations by an order of magnitude and has many applications in quality control and signal classification. As the observations sample Earth's climate sparsely in space and time, sophisticated statistical methods are used to map local signals to estimates of the full spatial field and its uncertainties. Much of the observed contraction of variability is shown to stem from decreased meridional temperature gradients due to amplified arctic warming in the northern hemisphere.

Short-term extremes are also contextualized with the low frequency variability inferred from paleoclimate observations and simulations. Spectral estimates used to measure variability on different time scales are shown to be surprisingly robust to unavoidable time-uncertainty present in all proxy records. Oxygen isotope records from Greenland that are widely used as temperature proxies therefore contain reliable signals of past climate variability on 1–60,000 year time scales, though the extent to which these reliably preserve temperature signals remains uncertain. In a further study we examine the validity of this temperature proxy using a set of global paleoclimate simulations with moisture source tracking, quantifying a seasonality bias that may explain the paleothermometer's damped response during glacial periods.

Greenland has experienced large changes since the last glacial with its summit warming by approximately 21°C, average accumulation rates tripling, and annual amplitudes of temperature and accumulation seemingly declining. The altered seasonal cycle of accumulation has been attributed to a combination of the large-scale dynamical response of the North Atlantic storm track to surface boundary conditions and the modulation of moisture availability due to changes in winter sea ice cover. Using atmospheric simulations of preindustrial and glacial climate, the contributions of these two mechanisms are evaluated. Estimates of moisture source footprints make it possible to distinguish between long-range transport related to the storm track and regional transport from the ocean surface near Greenland. It is found that the contribution of both mechanisms varies significantly with the background climate. With greater ice cover and the North Atlantic storm track locked to the topographically enhanced stationary wave during the glacial, seasonal migration of the sea ice edge becomes relatively important in controlling moisture availability. In contrast, the preindustrial simulation has relatively greater transient eddy activity and is less moisture limited by sea ice extent, so accumulation is more strongly related to synoptic variability in the North Atlantic. These results highlight how changes in atmospheric circulation and sea ice together explain the shifts in annual mean and seasonal moisture supply to Greenland. Also discussed are some implications of the inferred narrow source distribution of accumulation during the glacial for the interpretation of stable isotopes derived from the central Greenland ice cores.

Variations in extreme daily temperatures are explored in relation to changes in seasonal mean temperature using 1218 high-quality U.S. temperature stations spanning 1900–2012. Extreme temperatures are amplified (or damped) by as much as ±50% relative to changes in average temperature, depending on region, season, and whether daily minimum or maximum temperature is analyzed. The majority of this regional structure in amplification is shown to follow from regional variations in temperature distributions. More specifically, there exists a close relationship between departures from normality and the degree to which extreme changes are amplified relative to the mean. To distinguish between intraseasonal and interannual contributions to nonnormality and amplification, an additional procedure, referred to as z bootstrapping, is introduced that controls for changes in the mean and variance between years. Application of z bootstrapping indicates that amplification of winter extreme variations is generally consistent with nonnormal intraseasonal variability. Summer variability, in contrast, shows interannual variations in the spread of the temperature distribution related to changes in the mean, especially in the Midwest. Changes in midwestern temperature variability are qualitatively consistent with those expected from decreases in evapotranspiration and are strongly correlated with a measure of drought intensity. The identified patterns of interannual variations in means and extremes may serve as an analog for modes of variability that can be expected at longer time scales.

Historical observations of temperature underpin our ability to monitor Earth’s climate. We identify a pervasive issue in archived observations from surface stations, wherein the use of varying conventions for units and precision has led to distorted distributions of the data. Apart from the original precision being generally unknown, the majority of archived temperature data are found to be misaligned with the original measurements because of rounding on a Fahrenheit scale, conversion to Celsius, and re-rounding. Furthermore, we show that commonly used statistical methods including quantile regression are sensitive to the finite precision and to double-rounding of the data after unit conversion. To remedy these issues, we present a Hidden Markov Model that uses the differing frequencies of specific recorded values to recover the most likely original precision and units associated with each observation. This precision-decoding algorithm is used to infer the precision of the 644 million daily surface temperature observations in the Global Historical Climate Network database, providing more accurate values for the 63% of samples found to have been biased by double-rounding. The average absolute bias correction across the dataset is 0.018 ◦C, and the average inferred precision is 0.41 ◦C, even though data are archived at 0.1 ◦C precision. These results permit better inference of when record temperatures occurred, correction of rounding effects, and identification of inhomogeneities in surface temperature time series, amongst other applications. The precision-decoding algorithm is generally applicable to rounded observations–including surface pressure, humidity, precipitation, and other temperature data–thereby offering the potential to improve quality-control procedures for many datasets.

High temperature extremes during the growing season can reduce agricultural production. At the same time, agricultural practices can modify temperatures by altering the surface energy budget. Here we identify centennial trends towards more favourable growing conditions in the US Midwest, including cooler summer temperature extremes and increased precipitation, and investigate the origins of these shifts. Statistically significant correspondence is found between the cooling pattern and trends in cropland intensification, as well as with trends towards greater irrigated land over a small subset of the domain. Land conversion to cropland, often considered an important influence on historical temperatures, is not significantly associated with cooling. We suggest that agricultural intensification increases the potential for evapotranspiration, leading to cooler temperatures and contributing to increased precipitation. The tendency for greater evapotranspiration on hotter days is consistent with our finding that cooling trends are greatest for the highest temperature percentiles. Temperatures over rainfed croplands show no cooling trend during drought conditions, consistent with evapotranspiration requiring adequate soil moisture, and implying that modern drought events feature greater warming as baseline cooler temperatures revert to historically high extremes.

Nonnormal characteristics of geophysical time series are important determinants of extreme events and may provide insight into the underlying dynamics of a system. The structure of nonnormality in winter temperature is examined through the use of linear filtering of radiosonde temperature time series. Filtering either low or high frequencies generally suppresses what is otherwise statistically significant nonnormal variability in temperature. The structure of nonnormality is partly attributable to geometric relations between filtering and the appearance of skewness, kurtosis, and higher order moments in time series data, and partly attributable to the presence of nonnormal temperature variations at the highest resolved frequencies in the presence of atmospheric memory. A nonnormal autoregressive model and a multiplicative noise model are both consistent with the observed frequency structure of nonnormality. These results suggest that the generating mechanism for nonnormal variations does not necessarily act at the frequencies at which greatest nonnormality is observed.

Seasonal forecast models exhibit only modest skill in predicting extreme summer temperatures across the eastern US. Anomalies in sea surface temperature and monthly-resolution rainfall have, however, been correlated with hot days in the US, and seasonal persistence of these anomalies suggests potential for long-lead predictability. Here we present a clustering analysis of daily maximum summer temperatures from US weather stations between 1982–2015 and identify a region spanning most of the eastern US where hot weather events tend to occur synchronously. We then show that an evolving pattern of sea surface temperature anomalies, termed the Pacific Extreme Pattern, provides for skillful prediction of hot weather within this region as much as 50 days in advance. Skill is demonstrated using out-of-sample predictions between 1950 and 2015. Rainfall deficits over the eastern US are also associated with the occurrence of the Pacific Extreme Pattern and are demonstrated to offer complementary skill in predicting high temperatures. The Pacific Extreme Pattern appears to provide a cohesive framework for improving seasonal prediction of summer precipitation deficits and high temperature anomalies in the eastern US.