Predicting future climate at high spatial and temporal resolution.

Ilya M D Maclean

doi:10.1111/gcb.14876

Abstract

Most studies on the biological effects of future climatic changes rely on seasonally aggregated, coarse‐resolution data. Such data mask spatial and temporal variability in microclimate driven by terrain, wind and vegetation, and ultimately bear little resemblance to the conditions that organisms experience in the wild. Here, I present the methods for providing fine‐grained, hourly and daily estimates of current and future temperature and soil moisture over decadal timescales. Observed climate data and spatially coherent probabilistic projections of daily future weather were disaggregated to hourly and used to drive empirically calibrated physical models of thermal and hydrological microclimates. Mesoclimatic effects (cold‐air drainage, coastal exposure and elevation) were determined from the coarse‐resolution climate surfaces using thin‐plate spline models with coastal exposure and elevation as predictors. Differences between micro and mesoclimate temperatures were determined from terrain, vegetation and ground properties using energy balance equations. Soil moisture was computed in a thin upper layer and an underlying deeper layer, and the exchange of water between these layers was calculated using the van Genuchten equation. Code for processing the data and running the models is provided as a series of R packages. The methods were applied to the Lizard Peninsula, United Kingdom, to provide hourly estimates of temperature (100 m grid resolution over entire area, 1 m for a selected area) for the periods 1983–2017 and 2041–2049. Results indicated that there is a fine‐resolution variability in climatic changes, driven primarily by interactions between landscape features and decadal trends in weather conditions. High‐temporal resolution extremes in conditions under future climate change were predicted to be considerably less novel than the extremes estimated using seasonally aggregated variables. The study highlights the need to more accurately estimate the future climatic conditions experienced by organisms and equips biologists with the means to do so.

Highlights

Most studies on the climate biology are based on the climatic condi‐ tions above‐ground level seasonally averaged across 1 km2 or more (Potter, Woods, & Pincebourde, 2013)
Spatial variability in microclimate greatly exceeds the magnitude of climate change expected in the upcoming century, and ignoring this vari‐ ability leads to erroneous predictions of climate change impacts on species distributions (Gillingham, Huntley, Kunin, & Thomas, 2012; Lembrechts, Lenoir, et al, 2019; Lenoir, Hattab, & Pierre, 2017), population dynamics (Bennie et al, 2013) and behaviour (Blackshaw & Blackshaw, 1994)
While it is not possible to test how well the model performs under future conditions, the predictive capacity of the model was high, explaining over 90% of the variation in soil moisture and local temperature anomalies over the period in which validation was car‐ ried out

Summary

| INTRODUCTION

Most studies on the climate biology are based on the climatic condi‐ tions above‐ground level seasonally averaged across 1 km or more (Potter, Woods, & Pincebourde, 2013). While it is inherently impossible to predict the precise climate conditions experienced by an organism at some date and time in the distant future, reliable methods for generating synthetic time series of hourly or daily weather, using weather generators, are increasingly available (Ailliot, Allard, Monbet, & Naveau, 2015; Wilks & Wilby, 1999) Such ‘weather generators’ are capable of reproducing a wide set of climate statistics over a range of temporal scales, from the high‐frequency extremes to the low‐ frequency inter‐annual variability for future climate scenarios, as inferred from global climate models (Fatichi, Ivanov, & Caporali, 2011). These are compared to historic data generated for the 1983– 2017 period

| METHODS

Findings

| DISCUSSION