Estimating gambling venue catchments for impact assessment using a calibrated gravity model

Gambling using electronic gaming machines (EGMs) has emerged as a significant public health issue. While social impact assessments are required prior to the granting of new gaming machine licenses in Australia, there are a few established techniques for estimating the spatial distribution of a venue’s clientele. To this end, we calibrated a Huff model of gambling venue catchments based on a geocoded postal survey (n = 7040). We investigated the impact of different venue attractiveness measures, distance measures, distance decay functions, levels of spatial aggregation and venue types on model fit and results. We then compared model estimates for different behavioural subgroups. Our calibrated spatial model is a significant improvement on previously published models, increasing R² from 0.23 to 0.64. Venue catchments differ radically in size and intensity. As different population subgroups are attracted to different venues, there is no single best index of venue attractiveness applicable to all subpopulations. The calibrated Huff model represents a useful regulatory tool for predicting the extent and composition of gambling venue catchments. It may assist in decision-making with regard to new license applications and evaluating the impact of health interventions such as mandated reductions in EGM numbers. Our calibrated parameters may be used to improve model accuracy in other jurisdictions.     

Early Holocene water budget of the Nakuru-Elmenteita basin, Central Kenya Rift

The Nakuru-Elmenteita basin in the Central Kenya Rift, contains two shallow, alkaline lakes, Lake Nakuru (1770 m above sea level) and Lake Elmenteita (1786 m). Ancient shorelines and lake sediments at 1940 m suggest that these two lakes formed a single large and deep lake as a result of a wetter climate during the early Holocene. Here, we used a hydrological model to compare the precipitation–evaporation balance during the early Holocene to today. Assuming that the Nakuru-Elmenteita basin was hydrologically closed, as it is today, the most likely climate scenario includes a 45% increase in mean-annual precipitation, a 0.5°C decrease in air temperature, and an increase of 9% in cloud coverage from the modern values. Compared to the modeling results from other East African lake basins, this dramatic increase in precipitation seems to be unrealistic. Therefore, we propose a significant flow of water from the early Holocene Lake Naivasha in the south towards the Nakuru-Elmenteita basin to compensate the extremely negative hydrological budget of this basin. Since we did not find any field evidence for a surface connection, as often proposed during the last 70 years, the hydrological deficit of the Nakuru-Elmenteita basin could have also been compensated by a subsurface water exchange.     

The postglacial environmental development of Raffles Sø, East Greenland: inferences from a 10,000 year diatom record

A 341 cm long sediment sequence was recovered from the unofficially named Raffles Sø on Raffles Ø, outer Scoresby Sund region, East Greenland. The sediment sequence consists in the upper part (0–230 cm) of a stratified gyttja enriched in organic carbon and biogenic silica whereas the lower core part (235–341 cm) is composed of terrigenous, consolidated glacio-limnic sediments. ¹⁴C-AMS measurements indicate that the sediment sequence represents the entire Holocene lake history from 10,030 calibrated radiocarbon years.The geochemical parameters (opal, total organic carbon (TOC), total nitrogen (TN)) and the total diatom concentration show similar developments during the Holocene, and reflect changes in biological production and nutrient input into the lake. These records clearly reveal a broad Holocene TOC-opal-maximum interval between 5200 and 1800 cal. yrs BP.The diatom flora consisted of 66 taxa representing 20 genera but only seven taxa were abundant and, sometimes, these were monospecifically dominant during the Holocene. In the sediment core from Raffles Sø four successive stratigraphical zones can be distinguished. Accumulation of diatom valves began at 9900 cal. yrs BP with a Stephanodiscus minutulus (Kütz.) Cleve and Möller dominated assemblage (stratigraphic zone 1) followed by a diatom flora dominated by Cyclotella pseudostelligera Hustedt and, less frequently, by Diatoma tenuis Agardh (9400 until 5900 cal. yrs BP, zone 2). Cyclotella sp. A, a taxon which belongs to the Cyclotella rossii-comensis-tripartita-complex, was the dominant floral element between 5200 and 1800 cal. yrs BP (zone 3). From 1800 cal. yrs BP, the periphytic taxa Fragilaria capucina var. gracilis (Østr.) Hustedt and F. capucina var. rumpens (Kütz.) Lange-Bertalot attained highest relative abundances, also almost monospecifically (zone 4).The distribution and composition of the diatom assemblages in the sediment record from Raffles Sø probably reflect past variations in the extent of the lake-ice cover during the growing season. More or less ice-free conditions during summer may have prevailed during the early Holocene until ca. 1800 cal. yrs BP, which allowed growth of planktonic diatoms (Cyclotella taxa) in the pelagic lake region. From 1800 cal. yrs BP, colder conditions lead to a perennial lake-ice cover with a small ice-free moat in summer which favored the growth of periphytic, littoral species (Fragilaria capucina varieties).     

Climate variability and trends in downscaled high-resolution simulations and projections over Metropolitan France

In order to fulfill the society demand for climate information at the spatial scale allowing impact studies, long-term high-resolution climate simulations are produced, over an area covering metropolitan France. One of the major goals of this article is to investigate whether such simulations appropriately simulate the spatial and temporal variability of the current climate, using two simulation chains. These start from the global IPSL-CM4 climate model, using two regional models (LMDz and MM5) at moderate resolution (15–20 km), followed with a statistical downscaling method in order to reach a target resolution of 8 km. The statistical downscaling technique includes a non-parametric method that corrects the distribution by using high-resolution analyses over France. First the uncorrected simulations are evaluated against a set of high-resolution analyses, with a focus on temperature and precipitation. Uncorrected downscaled temperatures suffer from a cold bias that is present in the global model as well. Precipitations biases have a season- and model-dependent behavior. Dynamical models overestimate rainfall but with different patterns and amplitude, but both have underestimations in the South-Eastern area (Cevennes mountains) in winter. A variance decomposition shows that uncorrected simulations fairly well capture observed variances from inter-annual to high-frequency intra-seasonal time scales. After correction, distributions match with analyses by construction, but it is shown that spatial coherence, persistence properties of warm, cold and dry episodes also match to a certain extent. Another aim of the article is to describe the changes for future climate obtained using these simulations under Scenario A1B. Results are presented on the changes between current and mid-term future (2021–2050) averages and variability over France. Interestingly, even though the same global climate model is used at the boundaries, regional climate change responses from the two models significantly differ.     

Impact of climate change on the hydrogeology of two basins in northern France

This study presents an analysis of climate-change impacts on the water resources of two basins located in northern France, by integrating four sources of uncertainty: climate modelling, hydrological modelling, downscaling methods, and emission scenarios. The analysis focused on the evolution of the water budget, the river discharges and piezometric heads. Seven hydrological models were used, from lumped rainfall-discharge to distributed hydrogeological models, and led to quite different estimates of the water-balance components. One of the hydrological models, CLSM, was found to be unable to simulate the increased water stress and was, thus, considered as an outlier even though it gave fair results for the present day compared to observations. Although there were large differences in the results between the models, there was a marked tendency towards a decrease of the water resource in the rivers and aquifers (on average in 2050 about ?14 % and ?2.5 m, respectively), associated with global warming and a reduction in annual precipitation (on average in 2050 +2.1 K and ?3 %, respectively). The uncertainty associated to climate models was shown to clearly dominate, while the three others were about the same order of magnitude and 3–4 times lower. In terms of impact, the results found in this work are rather different from those obtained in a previous study, even though two of the hydrological models and one of the climate models were used in both studies. This emphasizes the need for a survey of the climatic-change impact on the water resource.