waters in drylands can also originate from recycling of groundwater via evapotranspiration and redistribution of water vapor in the upper soil layers. More than 50% of the total fog events in the NamibDesert were found to be non-ocean-derived or locally-generated (Kaseke et al., 2017). This suggests that this desert’s fog zone is potentially shifting from advection-dominated fog to radiation-dominated fog. This highlights the urgency of studies to determine the quantity and origin of non-rainfall waters in dryland ecosystems where rainfall is expected to decline due to climate change in coming years (Kaseke et al., 2017). Soilmoisture is considered as a critical component of earth system and plays an important role in land-atmosphere interactions (Eltahir, 1998). It can be used to understand the relationship of climate, soil and vegetation in dryland ecosystems (Li et al., 2016). In water-limited dryland ecosystems, various eco-hydrological processes are dependent on soilwater availability. Many factors like precipitation, evaporation, liquid water and water vapor flow influence soilmoisture near the surface in drylands, and most of these factors are strongly connected. In the NamibDesert, the first millimeters of surface soil might receive enough soilmoisture from non-rainfall waters like fog droplets and dew formation. Since soilwater content near the soil surface in drylands is often extremely low, water vapor transport plays a critical role in the overall water flux and availability (D'Odorico and Porporato, 2006; Saito et al., 2006).
The formation of NRW is aﬀected by climatic factors and soil surface properties together ( Kidron, 2000; Kidron and Temina, 2013; Wang et al., 2014 ). For example, more NRW forms in sunny days than that in cloudy days. Longer time of high relative humidity in nighttime and smaller di ﬀerence between air and dew-point are both beneﬁcial to higher NRW amount. Vapor pressure and atmospheric pressure can also aﬀect NRW by driving capillary eﬀect ( Agama and Berliner, 2006; Lan et al., 2010 ). Meanwhile, the presence of biocrusts signi ﬁcantly pro- motes the accumulation of NRW ( Liu et al., 2006; Zhang et al., 2009; Lan et al., 2010; Pan et al., 2010; Uclés et al., 2015 ), which is attributed to roughness, or soil texture ( Zhang et al., 2009; Pan et al., 2010; Zhuang and Zhao, 2014 ), exopolysaccharides and salinity ( Kosmas et al., 2001; Kidron et al., 2002; Heusinkveld et al., 2006; Chen et al., 2014; Colica et al., 2014; Wang et al., 2014 ). However, NRW derives from both atmosphere and soil ( Agam and Berliner, 2006; Lan et al., 2010 ), and the latter contains moisture stems from vapor and ca- pillarity. The portions from atmosphere and soil may be aﬀected by di ﬀerent properties, and the accumulation pattern of NRW should be the balanced result of biocrust traits and microclimate. Whereas, how diﬀerent biocrusts boost accumulation of NRW and how aﬀect diﬀerent parts of NRW are little known.
In non-rainfall periods, soil is also supplied with water from the atmosphere (Jacobson et al., 2015; Kidron et al., 2014; McHugh et al., 2015; Tomaszkiewicz et al., 2015). This is an effect that comes about due to the formation of dew, hoarfrost, soil-water vapour condensation and atmos- pheric water adsorption (Alishaev, 2013; Janik et al., 2014; Zhang et al., 2015). The importance of these processes in the water balance of the soil surface layer is demonstrated not only in arid (desert) areas, but also in humid regions (Cassity-Duffey and Cabrera, 2016). Infiltration in non- rainfall periods is defined as being the water flux from the atmosphere through the plane of soil surface (further on in the paper it is denoted with the symbol E R ). It is a physical
COMPARISON OF SOILMOISTURE OBSERVATIONS
Soilmoisture time‑series for all three sets of observations are plotted in Figure 2. Given the depths of the in situ measurements, and the characteristics of the COSMOS and ESA CCI measurements, the series derived from the soil probes is likely to represent the greatest depth of soil, and the satellite data the shallowest. This is consistent with the variability of the different time‑series with the TDT sensor data exhibiting the least daily variability, particularly during dry periods. As would be expected, given it represents only up to the top ~5cm of the soil, the ESA CCI data exhibits the greatest variability. Whilst there are obvious differences between the datasets the monthly to seasonal variability in the three series is in reasonable agreement, which suggests potential value in using satellite‑remote sensed soilmoisture to constrain spatio‑temporal estimates of groundwater recharge.
viii Figure 13: Location map of the study area in the NamibDesert gravel plains. The study area was divided into three sites (A, B and C). The distance between sites and their surface areas are depicted (Source: Google Earth; 6/27/2010). .................. 39 Figure 14: Photograph depicting Site A in the NamibDesert gravel plains. In May 2012, seasonal Stipagrostis sp., a common perennial grass in the region, was growing on the study site........................................................................................................................... 40 Figure 15: Photographs of the 100 m 2 Sites B and C study areas in the NamibDesert gravel plains in May 2012. These sites represented vegetation-free sites (Stipagrostis sp. was largely absent on the study sites) in this study and were located approximately 0.189 km apart............................................................................... 41 Figure 16: Equipment used in the sampling of desert surface soil in this study. 1 m 2 wired sampling grid (divided in 16 individual 25 x 25 cm quadrats) and a trowel is depicted. ....................................................................................................................... 42 Figure 17: Desertsoil total metagenomic DNA extraction. Lane 1: DNA molecular weight marker (Kappa), Lanes 2-8, soil metagenomic DNA. .................................................... 59 Figure 18: 16S rRNA gene PCR amplification from metagenomic DNA before (a) and after (b) optimization. a: Lane 1: DNA molecular marker (Kappa), Lanes 2-9, 16S rRNA gene amplification result, Lane 10, negative control. b: Lane 1: DNA molecular marker (Kappa),Lanes 2-4 16S rRNA gene amplification result, Lane 5, negative control. ..................................................................................................................... 60 Figure 19: Schematic diagram
The methodology for collecting pond depth data is de- scribed in Conly et al. (2004). The lowest bottom elevation in the wetland depression is used as the relative datum, and the geodetic elevations of these points have been determined for many of the wetlands. Measurements are made by wad- ing into the pond and using a measuring rod to measure water depth at monitoring markers (usually a metal T-bar installed deep into pond sediments to prevent heaving or movement). The measuring rod is attached to a 6 cm diameter circular base to prevent the rod from being pushed into the sediment. Shallow seasonal and ephemeral ponds require only a sin- gle marker. Deeper ponds that vary considerably in flooded area and depth have multiple markers installed at various el- evations to ensure a measurement can be made when mark- ers installed at lower elevations are flooded. Depth measure- ments are taken at the same time at multiple markers to en- sure markers are tied to the local datum. Point measurements at single markers (in smaller wetlands) are generally within 25 mm of those measured with conventional survey equip- ment and benchmarks. The accuracy in larger wetlands is considered to be within 50 mm (Conly et al., 2004). The year- to-year and seasonal variations in pond water levels are ap- parent from the long-term record (Fig. 4).
and runoff at year i, respectively ; DS i is the annual water
storage change at the watershed scale. The effects of water storage change on annual water balance have been consid- ered in several studies. Pike’s  functional form was based on the interannual variability of water balance for four watersheds in Malawi. The annual changes in ground- water storage were accounted for by constructing depletion curves under which the area was integrated to obtain a rela- tionship between ﬂow and storage left in the watershed at the end of the dry season. The annual storage change is negligible compared with precipitation and runoff in the four watersheds (Table 1 in the work of Pike ). Zhang et al.  found that Fu’s equation, one func- tional form of Budyko-type curves, performed poorly on estimating annual streamﬂow in some watersheds in Aus- tralia, and they explained that it might be because of the impact of watershed water storage, which could not be neglected at the annual scale. Donohue et al.  studied the annual water balance at 221 watersheds in Australia and found that the effect of nonsteady state conditions was an important source of variation at the annual scale and needed to be accounted for. During multiyear droughts, the annual storage change in the Murray Darling Basin can be up to twice the annual streamﬂow [Leblanc et al., 2009]. Flerchinger and Cooley  studied the water balance of the Upper Sheep Creek watershed, a 26-ha semiarid mountainous sub-basin within the Reynolds Creek experi- mental watershed in southwest Idaho, United States. Dur- ing 1985–1994, the minimum and maximum ratios of annual storage change (including soilmoisture and ground- water) to annual precipitation were 0.45 and 0.2, respec- tively, with the average absolute value of the ratios over the 10 yr being 0.16. The average ratio of annual runoff to annual precipitation (i.e., runoff coefﬁcient) was found to be 0.05. Thus, the annual storage carryover is signiﬁcant in this watershed. Milly and Dunne  accounted for the interannual storage change in the analysis of discharge var- iations for 175 large basins worldwide with a median area
Our conclusion that aspect is an important control on soilmoisture echoes the results of previous studies in NZ hill country (e.g. Bretherton et al., 2010; Lambert and Roberts, 1976). The mechanisms linking aspect with soilmoisture are varied. For example, Lambert and Roberts (1976) found complex interactions between air temperature, soil temper- ature and ET, driven by wind direction and aspect-induced radiation differences. They note that the specific heat ca- pacity of soil drops as it dries, leading to a positive feed- back cycle. In the Langs Gully catchment, the south-facing slopes are also steeper than the north-facing slopes. This is not obviously due to geological bedding – the main trend of syncline–anticline pairs in the wider Waipara catchment is northwest–southeast (transverse to catchment slopes), and in the immediate area of Langs Gully, known dip directions are highly variable. However, feedbacks are likely to exist be- tween slope angle, vegetation (denser shrub cover on south- facing slopes), soil depth (thinner on south-facing slopes) and downslope sediment transport. Shading by denser vege- tation and increased lateral flow are possible causes of the in- creased number of wetting events on the south-facing slope. Typical hydrological models do not account for aspect, but
Although desert soils support functionally important microbial communities that affect plant growth and influence many biogeochemical processes, the impact of future changes in precipitation patterns on the microbiota and their activities is largely unknown. We performed in-situ experiments to investigate the effect of simulated rainfall on bacterial communities associ- ated with the widespread perennial shrub, Rhazya stricta in Arabian desert soils. The bacterial communi- ty composition was distinct between three different soil compartments: surface biological crust, root- attached, and the broader rhizosphere. Simulated rain- fall had no significant effect on the overall bacterial community composition, but some population-level responses were observed, especially in soil crusts where Betaproteobacteria, Sphingobacteria, and Bacil- li became more abundant. Bacterial biomass in the nutrient-rich crust increased three-fold one week after watering, whereas it did not change in the rhizosphere, despite its much higher water retention. These find- ings indicate that between rainfall events, desert-soil microbial communities enter into stasis, with limited species turnover, and reactivate rapidly and relatively uniformly when water becomes available. However, microbiota in the crust, which was relatively enriched
Stream runoff is perhaps the most poorly represented process in ecohydrological stochastic soilmoisture models. Here we present a rainfall-runoff model with a new stochastic description of runoff linked to soilmoisture dynamics. We describe the rainfall-runoff system as the joint probability density function (PDF) of rainfall, soilmoisture and runoff forced by random, instantaneous jumps of rainfall. We develop a master equation for the soilmoisture PDF that accounts explicitly for a general state-dependent rainfall-runoff transformation. This framework is then used to derive the joint rainfall-runoff and soilmoisture-runoff PDFs. Runoff is initiated by a soilmoisture threshold and a linear progressive partitioning of rainfall based on the soilmoisture status. We explore the dependence of the PDFs on the rainfall occurrence PDF (homogeneous or state- dependent Poisson process) and the rainfall magnitude PDF (exponential or mixed-exponential distribution). We calibrate the model to 63 years of rainfall and runoff data from the Upper Little Tennessee watershed (USA) and show how the new model can reproduce the measured runoff PDF.
ment Tool (SWAT), with updating of multiple states and pa- rameters including runoff, soilmoisture and evapotranspira- tion. Camporese et al. (2009b) used EnKF in the CATHY (CATchment HYdrology) model with coupled surface and subsurface flow, to assimilate groundwater head and stream discharge. Rasmussen et al. (2015) assimilated the same vari- ables using the ensemble transform Kalman filter (ETKF) with the MIKE SHE model. Kurtz et al. (2014) jointly as- similated groundwater heads and groundwater temperatures with EnKF using both synthetic and real-world models. Shi et al. (2014) employed EnKF to assimilate multivariate hy- drological states in a small catchment modelled by the Flux- PIHM land surface model, with a focus on parameter esti- mation. Lee et al. (2011) used a variational assimilation ap- proach to assimilate streamflow and in situ soilmoisture, to correct the soilmoisture profiles within the HL-RDHM model. Ridler et al. (2014b) developed a generic DA frame- work that enables coupling of hydrological models with the OpenDA library (http://www.openda.org) using the OpenMI (Open Model Interface; Gregersen et al., 2007), and applied it with the MIKE SHE model. Han et al. (2015) developed an open-source multivariate DA framework (DasPy) for the Community Land Model. Although many multivariate DA platforms and applications have been reported, assimilating both soilmoisture and groundwater head in an integrated hy- drological model has not been studied in detail. Represent- ing two important hydrological variables, their observational values by assimilation in integrated hydrological models are explored in this study.
Hydrological, agricultural and water management applications have been also requesting soilmoisture datasets at a higher resolution for a long time now. In this context, downscaling methodologies have been developed to improve the spatial resolution of readily available passive microwave-derived soilmoisture data. In particular, the DisPATCh disaggregation scheme estimates the soilmoisture variability at high resolution within a low resolution pixel by relying on a self-calibrated evaporation model ( Merlin et al., 2013 ). The DisPATCh algorithm has been implemented and validated in several climatic regions such as Catalonia, Spain ( Merlin et al., 2012, 2013 ), Central Morocco ( Merlin et al., 2015 ), South-Eastern Australia ( Malbéteau et al., 2016 ) and two watersheds in the USA ( Molero et al., 2016 ). However, it has never been tested in the arid regions where the desert locust is likely to reproduce. In the microwave domain, active sensors (radars) achieve a spatial resolution much ﬁner than that of radiometers. Sentinel-1 A and B ( Torres et al., 2012 ) are providing C-band SAR (Synthetic Aperture Radar) data at a spatial resolution of about 20 m. Although backscatter data have potential to monitor SM ( Balenzano et al., 2013 ), there is currently no operational soilmoisture product at such ﬁne resolution. This is notably due to the di ﬃculty to model in time and over extended areas the impact of vegetation cover/structure and surface roughness on the backscatter signal ( Satalino et al., 2014 ), and thus the need for site-speci ﬁc calibration ( Zribi et al., 2011 ).
The soil wetness and soil type has a considerable role in studying the agricultural droughts. The amount of moisture actually present in the soil reflects the antecedent meteorological conditions; soil characteristics and the level of agronomic techniques at any given instant of time are in use in the region (Kulik, 1958). Taking soilmoisture content as a criterion Rodda (1965, 1969) made a study of droughts in southeast England. He calculated soilmoisture, drought index as a deviation value of available soilmoisture and the amounts of runoff and percolation.
Review of SoilMoisture Calculation Method
In accordance with the Habitats Directive, the Environment Agency is required to carry out an appropriate assessment when considering applications for abstraction licence renewals which may have an adverse impact on Sites of Special Scientific Interest (SSSIs). For some sites, one part of this assessment is the estimation of soilmoisture under various abstraction scenarios, and comparison with threshold values to ensure that abstraction from groundwater does not cause stress to vegetation on the SSSI. The estimation of soilmoisture content at sites above a shallow groundwater table has used a soilwater budget methodology, including calculations of capillary rise of groundwater from the water table to the root zone.
Abstract. The movement of subsurface water is mostly stud- ied at the pore scale and the Darcian scale, but the field and regional scales are of much larger societal interest. Volume- averaging has provided equations at these larger scales, but the required restrictions rendered them of little practical in- terest. Others hypothesized a direct connection at hydro- static equilibrium between the average matric potential of a subsurface body of water and the average pressure drop over the menisci in the soil pores. The link between the volume-averaged potential energy of subsurface water bodies and large-scale fluxes remains largely unexplored. This pa- per treats the effect of menisci on the potential energy of the water behind them in some detail, and discusses some field- scale effects of pore-scale processes. Then, various pub- lished expressions for volume-averaged subsurface water po- tentials are compared. The intrinsic phase average is deemed the best choice. The hypothesized relationship between aver- age matric potential and average meniscus curvature is found to be valid for unit gradient flow instead of hydrostatic equi- librium. Still, this restriction makes the relationship hold only for a specific depth range in the unsaturated zone un- der specific conditions, and certainly not for entire fields or catchments. In the groundwater, volume-averaged potential energy is of more use: for linearized, steady flows with flow lines that are parallel, radially diverging, and radially con- verging, proofs are derived for proportionality between av- eraged hydraulic potentials and fluxes towards open water at a fixed potential. For parallel flow, a simplified but relevant transient flow case also exhibits this proportionality.
The water table depth is the main indicator of the inten- sity of groundwater–soilmoisture coupling and consequently of how much memory the long timescales of variation of groundwater can induce in soilmoisture. The water table is linked to the unsaturated zone above by two-way fluxes: the downward gravitational flux and the capillary flux. The net flux is downward in the wet season and for some time af- terwards, when groundwater continues to be recharged, but upward capillary fluxes can dominate in the dry season and, if the water table is sufficiently shallow, groundwater will reach the root zone to meet surface ET demands. There is observational evidence from field experiments showing that groundwater can be one of the main sources of ecosystem ET in water-limited environments (e.g. Lubczynski, 2008; Liu et al., 2016) and that the groundwater table depth determines strong sensitivities of local rooting depths (Fan et al., 2017). In the Iberian Peninsula in particular, David et al. (2007) found that during the summer drought in a plot in southern Portugal, daily soilmoisture fluctuations in the top 1 m re- lated to transpiration could be attributed to groundwater via isotopic analysis. These authors estimated that up to 70 % of the evapotranspired water had its origin in groundwater over that area. Beyond experimental plots, observational evidence of the connection between groundwater and soilmoisture over a larger area is reported by Sutanudjaja et al. (2013), us- ing remote sensing soilmoisture products to predict ground- water heads in time and space over Germany and reproducing groundwater head fluctuations reasonably well, particularly in shallow water table areas, where soilmoisture dynamics are tightly connected to groundwater head positions.
Drought in recent years has highlighted the importance of maintaining a sustainable water resource. Improvements in irrigation management can significantly increase wa- ter use efficiency and crop productivity for Australian agriculture. Measurement of SoilMoisture Content (SMC) is essential for improving irrigation management. Ex- isting commercially-available SMC sensors require contact with the soil and measure only a single fixed point in a field. However, there can be significant spatial variability in soil properties and SMC within a field, and installation of multiple SMC sensors within a field is often not practical or economical. Non-contact methods reported in the literature for SMC estimation include satellite imagery of soil and plants. Satel- lite imagery approaches capture spectral bands in the visual, infrared and microwave wavelengths and then extract crop vigour to estimate SMC. However, this technology has a limited spatial resolution (30m 2 ) and temporal resolution (every 2-3 weeks). An alternative approach uses a ground-based camera that can be moved around the field on ground-based or aerial vehicles as required, providing high spatial and temporal res- olution SMC estimation. A camera-based estimation system has been developed. Red and near infrared images of plants are processed using MATLAB R Image Processing
In this paper, we discuss the configuration and perfor- mance of a new station constructed in 2012 at a near-coastal background site at Gobabeb (GAW station ID “NMB”), Namibia, here referred to as the NamibDesert Atmo- spheric Observatory (NDAO). Surface flask samples have been taken near Gobabeb since 1997 as part of the US National Oceanic and Atmospheric Administration Earth System Research Laboratory Global Monitoring Division (NOAA ESRL GMD) Carbon Cycle Cooperative Global Air Sampling Network. The NMB site is ∼ 3 km from NDAO. The goals of this project are to expand the ground-based sta- tion network in Africa and to use the time series as a top- down perspective on regional biogeochemical cycling and surface–atmosphere exchange of GHGs. The particular fo- cus of the project at a regional level is on the influence of biomass burning and coastal upwelling on GHG budgets. As the site receives air largely free from anthropogenic influ- ences, it is representative of both the terrestrial and the ma- rine background (Fig. 1). The main quantities measured at the observatory are carbon dioxide, methane, nitrous oxide, carbon monoxide, and atmospheric oxygen.
1997 ) located in an experimental plot of 1 ha (see Fig. 1 ). The average of the six sensors at different depths is used as ‘observed’ soilmoisture data here in after. Rainfall is measured by four raingauge stations located within (or close to) the catchment while the water level, converted in flow discharge through a reliable rating curve, is continuously monitored at the outlet (see Fig. 1 ). Figure 2 shows temporal variation of measured soilmoisture data, along with precipitation and discharge for the Colorso catchment. As it can be seen, soilmoisture at 10 cm depth shows a rapid rise during heavy rainfall, followed by a recession which differs according to the period of the year and to the meteorological conditions. The sensors at 20 and 40 cm depth show a delayed response to rainfall and also saturated conditions during very wet periods.