Using active learning for monitoring networks design: The example of wind power plants sites evaluation

Locating new wind farms is of crucial importance for energy policies of the next decade. To select the new location, an accurate picture of the wind fields is necessary. However, characterizing wind fields is a difficult task, since the phenomenon is highly nonlinear and related to complex topographical features. In this paper, we propose both a nonparametric model to estimate wind speed at different time instants and a procedure to discover underrepresented topographic conditions, where new measuring stations could be added. Compared to space filling techniques, this last approach privileges optimization of the output space, thus locating new potential measuring sites through the uncertainty of the model itself.

Density-dependent pollen limitation and reproductive assurance in a wind-pollinated herb with contrasting sexual systems

1. Wind pollination is thought to have evolved in response to selection for mechanisms to promote pollination success, when animal pollinators become scarce or unreliable. We might thus expect wind-pollinated plants to be less prone to pollen limitation than their insect-pollinated counterparts. Yet, if pollen loads on stigmas of wind-pollinated species decline with distance from pollen donors, seed set might nevertheless be pollen-limited in populations of plants that cannot self-fertilize their progeny, but not in self-compatible hermaphroditic populations.2. Here, we test this hypothesis by comparing pollen limitation between dioecious and hermaphroditic (monoecious) populations of the wind-pollinated herb Mercurialis annua.3. In natural populations, seed set was pollen-limited in low-density patches of dioecious, but not hermaphroditic, M. annua, a finding consistent with patterns of distance-dependent seed set by females in an experimental array. Nevertheless, seed set was incomplete in both dioecious and hermaphroditic populations, even at high local densities. Further, both factors limited the seed set of females and hermaphrodites, after we manipulated pollen and resource availability in a common garden experiment.4. Synthesis. Our results are consistent with the idea that pollen limitation plays a role in the evolution of combined vs. separate sexes in M. annua. Taken together, they point to the potential importance of pollen transfer between flowers on the same plant (geitonogamy) by wind as a mechanism of reproductive assurance and to the dual roles played by pollen and resource availability in limiting seed set. Thus, seed set can be pollen-limited in sparse populations of a wind-pollinated species, where mates are rare or absent, having potentially important demographic and evolutionary implications.

Sexual dimorphism in androdioecious Mercurialis annua, a wind-pollinated herb

Gender-dimorphic species often display a degree of sexual dimorphism in terms of life-history traits, yet little is known about dimorphism in androdioecious plants. Here we investigate sexual dimorphism in an androdioecious population of the wind-pollinated herb Mercurialis annua by comparing the resource allocation strategies of males and hermaphrodites grown under different nutrient-availability and competitive regimes. We found that males displayed smaller aboveground vegetative sizes than did hermaphrodites, but neither soil nutrient availability nor competition had a strong independent effect on their relative sizes. Plants adjusted their relative reproductive investment in response to nutrient availability. Specifically, hermaphrodites increased their reproductive allocation when growing in poor soils, whereas males displayed the opposite response. Finally, hermaphrodites were strongly female biased in their sex allocation, and this was more pronounced in nutrient-poor soils. To conclude, sexual dimorphism in androdioecious M. annua shares many features with dioecious and gynodioecious species, particularly wind-pollinated herbs. However, the direction of sex-allocation reaction norms displayed by hermaphrodites of M. annua differs from that documented for several insect-pollinated gynodioecious species, hinting at the importance of either the pollination mode or the sexual system as a context of selection shaping the reproductive strategy of plants with both male and female functions.

Learning wind fields with multiple kernels

This paper presents multiple kernel learning (MKL) regression as an exploratory spatial data analysis and modelling tool. The MKL approach is introduced as an extension of support vector regression, where MKL uses dedicated kernels to divide a given task into sub-problems and to treat them separately in an effective way. It provides better interpretability to non-linear robust kernel regression at the cost of a more complex numerical optimization. In particular, we investigate the use of MKL as a tool that allows us to avoid using ad-hoc topographic indices as covariables in statistical models in complex terrains. Instead, MKL learns these relationships from the data in a non-parametric fashion. A study on data simulated from real terrain features confirms the ability of MKL to enhance the interpretability of data-driven models and to aid feature selection without degrading predictive performances. Here we examine the stability of the MKL algorithm with respect to the number of training data samples and to the presence of noise. The results of a real case study are also presented, where MKL is able to exploit a large set of terrain features computed at multiple spatial scales, when predicting mean wind speed in an Alpine region.

Multiple Kernel Learning of Environmental Data. Case study: Analysis and Mapping of Wind Fields

The paper presents the Multiple Kernel Learning (MKL) approach as a modelling and data exploratory tool and applies it to the problem of wind speed mapping. Support Vector Regression (SVR) is used to predict spatial variations of the mean wind speed from terrain features (slopes, terrain curvature, directional derivatives) generated at different spatial scales. Multiple Kernel Learning is applied to learn kernels for individual features and thematic feature subsets, both in the context of feature selection and optimal parameters determination. An empirical study on real-life data confirms the usefulness of MKL as a tool that enhances the interpretability of data-driven models.

Spatial prediction of monthly wind speeds in complex terrain with adaptive general regression neural networks

This paper presents the general regression neural networks (GRNN) as a nonlinear regression method for the interpolation of monthly wind speeds in complex Alpine orography. GRNN is trained using data coming from Swiss meteorological networks to learn the statistical relationship between topographic features and wind speed. The terrain convexity, slope and exposure are considered by extracting features from the digital elevation model at different spatial scales using specialised convolution filters. A database of gridded monthly wind speeds is then constructed by applying GRNN in prediction mode during the period 1968-2008. This study demonstrates that using topographic features as inputs in GRNN significantly reduces cross-validation errors with respect to low-dimensional models integrating only geographical coordinates and terrain height for the interpolation of wind speed. The spatial predictability of wind speed is found to be lower in summer than in winter due to more complex and weaker wind-topography relationships. The relevance of these relationships is studied using an adaptive version of the GRNN algorithm which allows to select the useful terrain features by eliminating the noisy ones. This research provides a framework for extending the low-dimensional interpolation models to high-dimensional spaces by integrating additional features accounting for the topographic conditions at multiple spatial scales. Copyright (c) 2012 Royal Meteorological Society.

Sexual dimorphism in a dioecious population of the wind-pollinated herb Mercurialis annua: the interactive effects of resource availability and competition.

Background and Aims Male-biased sex allocation commonly occurs in wind-pollinated hermaphroditic plants, and is often positively associated with size, notably in terms of height. Currently, it is not well established whether a corresponding pattern holds for dioecious plants: do males of wind-pollinated species exhibit greater reproductive allocation than females? Here, sexual dimorphism is investigated in terms of life history trade-offs in a dioecious population of the wind-pollinated ruderal herb Mercurialis annua.Methods The allocation strategies of males and females grown under different soil nutrient availability and competitive (i.e. no, male or female competitor) regimes were compared.Key Results Male reproductive allocation increased disproportionately with biomass, and was greater than that of females when grown in rich soils. Sexual morphs differentially adjusted their reproductive allocation in response to local environmental conditions. In particular, males reduced their reproductive allocation in poor soils, whereas females increased theirs, especially when competing with another female rather than growing alone. Finally, males displayed smaller above-ground vegetative sizes than females, but neither nutrient availability nor competition had a strong independent effect on relative size disparities between the sexes.Conclusions Selection appears to favour plasticity in reproductive allocation in dioecious M. annua, thereby maintaining a relatively constant size hierarchy between sexual morphs. In common with other dioecious species, there seems to be little divergence in the niches occupied by males and females of M. annua.