On the need to test hydrological models under changing conditions

Abstract

The ability of hydrological models to deal with changing conditions should not be taken for granted: it is an unfortunate but well-known problem of hydrology that the model structure and/or parameters optimized for certain conditions may not be transferable in time. Consequently, it is essential that, for application under changing conditions (e.g. in climate change impact studies), models be thoroughly assessed for their extrapolation capacity using adequate protocols. This editorial provides an overview of the Special Issue of Hydrological Sciences Journal compiled after a workshop on this theme held during the General Assembly of the International Association of Hydrological Sciences (IAHS) in Gothenburg (Sweden) in 2013. The Workshop participants had been invited to apply a calibration and evaluation protocol to their own models on a given set of changing basins. The results show that this protocol is an appropriate and instructive way of assessing the suitability of hydrological models to be applied under changing conditions. This special issue also includes papers following alternative testing methodologies, as well as an opinion paper on the definition of non-stationarity.     

Optical long baseline intensity interferometry: prospects for stellar physics

More than sixty years after the first intensity correlation experiments by Hanbury Brown and Twiss, there is renewed interest for intensity interferometry techniques for high angular resolution studies of celestial sources. We report on a successful attempt to measure the bunching peak in the intensity correlation function for bright stellar sources with 1 meter telescopes (I2C project). We propose further improvements of our preliminary experiments of spatial interferometry between two 1 m telescopes, and discuss the possibility to export our method to existing large arrays of telescopes.     

Western boundary upwelling dynamics off Oman

Despite its climatic and ecosystemic significance, the coastal upwelling that takes place off Oman is not well understood. A primitive-equation, regional model forced by climatological wind stress is used to investigate its dynamics and to compare it with the better-known Eastern Boundary Upwellings (EBUs). The solution compares favorably with existing observations, simulating well the seasonal cycles of thermal structure, surface circulation (mean and turbulent), and sea-surface temperature (SST). There is a 1.5-month lag between the maximum of the upwelling-favorable wind-stress-curl forcing and the oceanic response (minima in sea-surface height and SST), which we attribute to onshore-propagating Rossby waves. A southwestward-flowing undercurrent (opposite to the direction of the near-surface flow) is also simulated with a core depth of 1000 m, much deeper than found in EBUs (150–200 m). An EKE budget reveals that, in contrast to EBUs, the upwelling jet is more prone to barotropic than baroclinic instability and the contribution of locally-generated instabilities to EKE is higher by an order of magnitude. Advection and redistribution of EKE by standing mesoscale features also play a significant role in EKE budget.     

Shifting zonal patterns of the southern boreal forest in eastern Canada associated with changing fire regime during the Holocene

This research aims at uncovering the stand-scale Holocene fire history of balsam fir forest stands from two bioclimatic zones of the boreal forest and assessing the existence of a sub-continental shift in past fire activity that could have triggered a change in the Holocene zonal pattern. In eastern Canada, the extant closed-crown boreal forest corresponds to two ecological regions separated along 49°N, the northern black spruce zone and the southern balsam fir zone. We sampled balsam fir stands from the southern fir zone (n = 7) and among the northernmost patches of fir forest located far beyond the fir zone boundary, into the spruce zone (n = 6). Macrofossil analysis of charcoal in mineral soils was used to reconstruct both the stand-scale and regional Holocene fire histories. Data were interpreted in the context of published palaeoecological evidence. Stands of the balsam fir zone were submitted to recurrent fire disturbances between c. 9000 and 5000 cal. yr B.P. Local fire histories suggested that four sites within the fir zone escaped fire during the Holocene. Such fire protected sites allowed the continuous maintenance of the balsam fir forest in the southern boreal landscape. Stands of the spruce zone have been affected by recurrent fires from 5000 cal. yr B.P. to present. Local fire histories indicated that no site escaped fire in this zone. Published palaeoecological data suggested that balsam fir migrated to its current northern limit sometime between 7300 and 6200 cal. yr B.P. A change of the fire regime 5000 years ago caused the regional decline of an historical northern balsam fir forest and its replacement by black spruce forest. The consequence was a sub-continental reshuffling of the fir and spruce zones within the closed-crown boreal forest. The macrofossil analysis of charcoal in mineral soils was instrumental to the reconstruction of stand-scale Holocene fire history at sites where no other in situ fire proxies were available.     

Spectral analysis of mean flow and turbulence forced by waves in a horizontally homogeneous zone of the Iroise sea

The 1D version of the Model for Applications at Regional Scale is used to parameterize the effects of sea surface waves in 2D in a horizontally homogeneous offshore zone of the Iroise sea. Here we present the first simulation of the Iroise sea including sea surface waves forcing, and more generally, the first study of a boundary layer including the Hasselmann force with a tidal wave. We use a single equation turbulence closure based on a non-local diagnosis for energetic and dissipation length scales. The turbulent energy flux at the surface due to whitecaps and the Hasselmann force induced by Stokes drift are assessed using the whole sea surface waves spectrum given by the Wave Watch Third generation model. The ability of the parameterization to reproduce surface currents over a period of 1 year (2007) is tested with high frequency radar using spectral and time-frequency analysis. One problem with 1D modelling, corresponding to overestimation of current oscillating at inertial frequency is illustrated by comparing 1D and 3D simulations. We found an overall improvement by including the Hasselmann force mainly within the bandwidth of less than one cycle per day to one cycle per day for surface currents. Turbulence is induced by whitecaps decaying rapidly below the ocean surface but the mixed layer below 40 m is deeper due to waves breaking on the sea surface.     

A high-precision digital astrogeodetic traverse in an area of steep geoid gradients close to the coast of Perth,Western Australia

We present results from a new vertical deflection (VD) traverse observed in Perth, Western Australia, which is the first of its kind in the Southern Hemisphere. A digital astrogeodetic QDaedalus instrument was deployed to measure VDs with \({\sim }\)0.2\(''\) precision at 39 benchmarks with a \({{\sim }}1~\hbox {km}\) spacing. For the conversion of VDs to quasigeoid height differences, the method of astronomical–topographical levelling was applied, based on topographical information from the Shuttle Radar Topography Mission. The astronomical quasigeoid heights are in 20–30 mm (RMS) agreement with three independent gravimetric quasigeoid models, and the astrogeodetic VDs agree to 0.2–0.3\(''\) (north–south) and 0.6–0.9\(''\) (east–west) RMS. Tilt-like biases of \({\sim }1\,\,\hbox {mm}\) over \({\sim }1\,\,\hbox {km}\) are present for all quasigeoid models within \({\sim }20\,\,\hbox {km}\) of the coastline, suggesting inconsistencies in the coastal zone gravity data. The VD campaign in Perth was designed as a low-cost effort, possibly allowing replication in other Southern Hemisphere countries (e.g., Asia, Africa, South America and Antarctica), where VD data are particularly scarce.     

Beach recovery from extreme storm activity during the 2013–14 winter along the Atlantic coast of Europe

The storm sequence of the 2013–14 winter left many beaches along the Atlantic coast of Europe in their most eroded state for decades. Understanding how beaches recover from such extreme events is essential for coastal managers, especially in light of potential regional increases in storminess due to climate change. Here we analyse a unique dataset of decadal beach morphological changes along the west coast of Europe to investigate the post-2013–14 winter recovery. We show that the recovery signature is site specific and multi-annual, with one studied beach fully recovered after 2 years, and the others only partially recovered after 4 years. During the recovery phase, winter waves primarily control the timescales of beach recovery, as energetic winter conditions stall the recovery process whereas moderate winter conditions accelerate it. This inter-annual variability is well correlated with climate indices. On exposed beaches, an equilibrium model showed significant skill in reproducing the post-storm recovery and thus can be used to investigate the recovery process in more detail. © 2018 John Wiley & Sons, Ltd.     

D/O and D/H ratios in the local interstellar medium from FUSE observations

Since HI, OI, and DI have nearly the same ionization potential, the deuterium-to-oxygen ratio (D/O) is an important tracer of the D/H ratio and of its putative spatial variations. D/O is indeed very sensitive to astration, both because of deuterium destruction and oxygen production. Moreover, D/O measurements are less subject to systematic errors than D/H.A survey of D/O in the interstellar medium is performed by the Deuterium Working Group of the FUSE PI Team. The first results show that D/O is a better D/H proxy than D/N, and that D/O is homogeneous in the local interstellar medium, at least within about 100pc. The stability of D/O appears as a strong argument which supports both D/H and O/H spatial stability in the local interstellar medium.     

Source to sink relations between the Tian Shan and Junggar Basin (northwest China) from Late Palaeozoic to Quaternary: evidence from detrital U‐Pb zircon geochronology

The tectonic evolution of the Tian Shan, as for most ranges in continental Asia is dominated by north‐south compression since the Cenozoic India‐Asia collision. However, precollision governing tectonic processes remain enigmatic. An excellent record is provided by thick Palaeozoic – Cenozoic lacustrine to fluvial depositional sequences that are well preserved in the southern margin of the Junggar Basin and exposed along a foreland basin associated to the Late Cenozoic rejuvenation of the Tian Shan ranges. U/Pb (LA‐ICP‐MS) dating of detrital zircons from 14 sandstone samples from a continuous series ranging in age from latest Palaeozoic to Quaternary is used to investigate changes in sediment provenance through time and to correlate them with major tectonic phases in the range. Samples were systematically collected along two nearby sections in the foreland basin. The results show that the detrital zircons are mostly magmatic in origin, with some minor input from metamorphic zircons. The U‐Pb detrital zircon ages range widely from 127 to 2856 Ma and can be divided into four main groups: 127–197 (sub‐peak at 159 Ma), 250–379 (sub‐peak at 318 Ma), 381–538 (sub‐peak at 406 Ma) and 543–2856 Ma (sub‐peak at 912 Ma). These groups indicate that the zircons were largely derived from the Tian Shan area to the south since a Late Carboniferous basin initiation. The provenance and basin‐range pattern evolution of the southern margin of Junggar Basin can be generally divided into four stages: (1) Late Carboniferous – Early Triassic basin evolution in a half‐graben or post‐orogenic extensional context; (2) From Middle Triassic to Upper Jurassic times, the southern Junggar became a passively subsiding basin until (3) being inverted during Lower Cretaceous – Palaeogene; (4) During the Neogene, a piedmont developed along the northern margin of the North Tian Shan block and Junggar Basin became a true foreland basin.     

Routes of Agulhas rings in the southeastern Cape Basin

Using weekly sea surface height data, Agulhas rings from the period October 1992 to December 2006 are detected and tracked, from their formation dates and throughout the Cape Basin. While 102 of them formed at the Agulhas Current retroflection, their subsequent subdivisions and junctions led to 199 trajectories. The rings geographical probability of presence shows two maxima. One, related to numerous ring passages, lies in the submarine bight formed by the Erica seamount, the Schmitt-Ott seamount, and the northeastern tip of the Agulhas Ridge. The other one, to be ascribed to topographic blocking of the eddies, is southeast of the latter obstacle. On the basis of topographic effects three routes for Agulhas rings are distinguished, a Northern route for rings that enter the south-Atlantic northeast of the Erica seamount, a Central one for those passing westward between this seamount and the tip of the Agulhas Ridge, and a Southern one farther south. Despite its bathymetric obstacles, the central route is the dominant one, both in terms of percentage of eddy crossings at its definition segment, and in terms of conveyed volume transport. Specific behaviours of rings along each route are described, referring to observations in previous studies. Some rings from the Northern route interact with the flow regime of the South African continental slope. The southernmost trajectories of the Central route are thought to settle the location of the climatological Subtropical Front in that region. The rings of the Southern route experience important core property alteration as they transit through the subantarctic domain.