Looking back and moving forward: Health geography in Aotearoa and beyond

Health geography in Aotearoa New Zealand (NZ) has had a major impact on international sub‐disciplinary debates. It is timely to celebrate our domestic health geography achievements and alert ourselves to the challenges ahead. This reflection article provides a brief overview of Aotearoa NZ's maturing health geography community and includes the reflections of a panel discussion held during the bi‐annual NZ Geographical Society conference in 2016. We argue that the eclectism of health geography in NZ and beyond is both the sub‐discipline's major strength and challenge.     

Data mining methods for hydroclimatic forecasting

Skillful streamflow forecasts at seasonal lead times may be useful to water managers seeking to provide reliable water supplies and maximize hydrosystem benefits. In this study, a class of data mining techniques, known as tree-structured models, is investigated to address the nonlinear dynamics of climate teleconnections and screen promising probabilistic streamflow forecast models for river–reservoir systems. In a case study of the Lower Colorado River system in central Texas, a number of potential predictors are evaluated for forecasting seasonal streamflow, including large-scale climate indices related to the El Niño–Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), and others. Results show that the tree-structured models can effectively capture the nonlinear features hidden in the data. Skill scores of probabilistic forecasts generated by both classification trees and logistic regression trees indicate that seasonal inflows throughout the system can be predicted with sufficient accuracy to improve water management, especially in the winter and spring seasons in central Texas.     

Model support for forcing of the 8.2 ka event by meltwater from the Hudson Bay ice dome

Previous model experiments of the 8.2 ka event forced by the drainage of Lake Agassiz often do not produce climate anomalies as long as those inferred from proxies. In addition to the Agassiz forcing, there is new evidence for significant amounts of freshwater entering the ocean at 8.2 ka from the disintegration of the Laurentide ice sheet (LIS). We use the Community Climate System Model version 3 (CCSM3) to test the contribution of this additional meltwater flux. Similar to previous model experiments, we find that the estimated freshwater forcing from Lake Agassiz is capable of sustaining ocean and climate anomalies for only two to three decades, much shorter than the event duration of ~150 years in proxies. Using new estimates of the LIS freshwater flux (~0.13 Sv for 100 years) from the collapse of the Hudson Bay ice dome in addition to the Agassiz drainage, the CCSM3 generates climate anomalies with a magnitude and duration that match within error those from proxies. This result is insensitive to the duration of freshwater release, a major uncertainty, if the total volume remains the same. An analysis of the modeled North Atlantic freshwater budget indicates that the Agassiz drainage is rapidly transported out of the North Atlantic while the LIS contribution generates longer-lasting freshwater anomalies that are also subject to recirculation by the subtropical gyre back into the North Atlantic. Thus, the meltwater flux originating from the LIS appears to be more important than the Agassiz drainage in generating 8.2 ka climate anomalies and is one way to reconcile some model-data discrepancies.     

Wavelet Analysis and Filtering to Identify Dominant Orientations of Permeability Anisotropy

An accurate representation of permeability anisotropy is needed to model the rate and direction of groundwater flow correctly. We develop a wavelet analysis technique that can be used to characterize principal directions of anisotropy in both stationary and non-stationary permeability fields. Wavelet analysis involves the integral transform of a field using a wavelet as a kernel. The wavelet is shifted, scaled, and rotated to analyze different locations, sizes, and orientations of the field. The wavelet variance is used to identify scales and orientations that dominate the field. If the field is non-stationary, such that different zones of the field are characterized by different dominant scales or orientations, the wavelet variance can identify all dominant scales and orientations if they are distinct. If the dominant scales and orientations of different zones are similar, the wavelet variance identifies only the dominant scale and orientation of the primary zone. In this paper, we present a combined wavelet analysis and filtering approach to identify all dominant scales and orientations in a non-stationary permeability field. We apply the method to permeability data obtained in the laboratory from the Massillon sandstone.     

Variability and predictability of Antarctic krill swarm structure

Swarming is a fundamental part of the life of Euphausia superba, yet we still know very little about what drives the considerable variability in swarm shape, size and biomass. We examined swarms across the Scotia Sea in January and February 2003 using a Simrad EK60 (38 and 120 kHz) echosounder, concurrent with net sampling. The acoustic data were analysed through applying a swarm-identification algorithm and then filtering out all non-krill targets. The area, length, height, depth, packing-concentration and inter-swarm distance of 4525 swarms was derived by this method. Hierarchical clustering revealed 2 principal swarm types, which differed in both their dimensions and packing-concentrations. Type 1 swarms were generally small (<50 m long) and were not very tightly packed (<10 ind. m⁻³), whereas type 2 swarms were an order of magnitude larger and had packing concentrations up to 10 times greater. Further sub-divisions of these types identified small and standard swarms within the type 1 group and large and superswarms within the type 2 group. A minor group (swarm type 3) was also found, containing swarms that were isolated (>100 km away from the next swarm). The distribution of swarm types over the survey grid was examined with respect to a number of potential explanatory variables describing both the environment and the internal-state of krill (namely maturity, body length, body condition). Most variables were spatially averaged over scales of 100 km and so mainly had a mesoscale perspective. The exception was the level of light (photosynthetically active radiation (PAR)) for which measurements were specific to each swarm. A binary logistic model was constructed from four variables found to have significant explanatory power (P<0.05): surface fluorescence, PAR, krill maturity and krill body length. Larger (type 2) swarms were more commonly found during nighttime or when it was overcast during the day, when surface fluorescence was low, and when the krill were small and immature. A strong pattern of diel vertical migration was not observed although the larger and denser swarms tended to occur more often at night than during the day. The vast majority of krill were contained within a minor fraction of the total number of swarms. These krill-rich swarms were more common in areas dominated by small and immature krill. We propose that, at the mesoscale level, the structure of swarms switches from being predominantly large and tightly packed to smaller and more diffuse as krill grow and mature. This pattern is further modulated according to feeding conditions and then level of light.     

Determination of Cepheid parameters by light-curve template fitting

We describe techniques to characterize the light curves of regular variable stars by applying principal component analysis (PCA) to a training set of high-quality data, and to fit the resulting light-curve templates to sparse and noisy photometry to obtain parameters such as periods, mean magnitudes etc. The PCA approach allows us to efficiently represent the multiband light-curve shapes (LCSs) of each variable, and hence quantitatively describe the average behaviour of the sample as a smoothly varying function of period, and also the range of variation around this average.
In this paper we focus particularly on the utility of such methods for analysing Hubble Space Telescope ( HST ) Cepheid photometry, and present simulations which illustrate the advantages of our PCA template-fitting approach. These are: accurate parameter determination, including LCS information; simultaneous fitting to multiple passbands; quantitative error analysis; objective rejection of variables with non-Cepheid-like light curves or those with potential period aliases.
We also use PCA to confirm that Cepheid LCSs are systematically different (at the same period) between the Milky Way and the Large and Small Magellanic Clouds, and consider whether LCS might therefore be used to estimate the mean metallicities of Cepheid samples, thus allowing metallicity corrections to be applied to derived distance estimates.     

Comparison of solutions to the thirteen-moment and standard transport equations for low speed thermal proton flows

We have compared solutions obtained from the general 13-moment system of transport equations with those obtained from the standard collision-dominated transport equations for conditions corresponding to low speed thermal proton flow in the topside ionosphere in the vicinity of the plasmapause. In general, the solutions obtained from the 13-moment system of equations, which allows for different species temperatures parallel and perpendicular to the geomagnetic field and non-classical heat flows, are different from those obtained from the standard transport equations, which account for isotropic temperatures and classical collision-dominated heat flows. Within the plasmasphere, where the electron density is high, the differences between the 13-moment and standard solutions are typically small. However, outside the plasmasphere where the electron density is lower and in the ionosphere above SAR-arcs, where substantial electron and proton heat flows occur, there can be significant differences between the 13-moment and standard solutions. Generally, the differences are much larger for the protons than for the electrons. Our 13-moment solutions indicate that the proton and electron distributions are anisotropic with the difference between parallel and perpendicular temperatures approaching 4000 K for the protons and 2500 K for the electrons in the ionosphere above SAR-arcs. Also, above SAR-arcs the 13-moment heat flow equations yield proton heat flows as much as a factor of 10 lower and electron heat flows as much as a factor of 2 lower than those predicted by the classical collision-dominated heat flow expressions for the same boundary conditions.