Climate change and urbanization can increase pressures on groundwater resources, but little is known about how groundwater quality will change. Here, we use a global synthesis (n = 9,404) to reveal ...the drivers of dissolved organic carbon (DOC), which is an important component of water chemistry and substrate for microorganisms that control biogeochemical reactions. Dissolved inorganic chemistry, local climate and land use explained ~ 31% of observed variability in groundwater DOC, whilst aquifer age explained an additional 16%. We identify a 19% increase in DOC associated with urban land cover. We predict major groundwater DOC increases following changes in precipitation and temperature in key areas relying on groundwater. Climate change and conversion of natural or agricultural areas to urban areas will decrease groundwater quality and increase water treatment costs, compounding existing constraints on groundwater resources.
Groundwater in sub-Saharan Africa supports livelihoods and poverty alleviation
, maintains vital ecosystems, and strongly influences terrestrial water and energy budgets
. Yet the hydrological ...processes that govern groundwater recharge and sustainability-and their sensitivity to climatic variability-are poorly constrained
. Given the absence of firm observational constraints, it remains to be seen whether model-based projections of decreased water resources in dry parts of the region
are justified. Here we show, through analysis of multidecadal groundwater hydrographs across sub-Saharan Africa, that levels of aridity dictate the predominant recharge processes, whereas local hydrogeology influences the type and sensitivity of precipitation-recharge relationships. Recharge in some humid locations varies by as little as five per cent (by coefficient of variation) across a wide range of annual precipitation values. Other regions, by contrast, show roughly linear precipitation-recharge relationships, with precipitation thresholds (of roughly ten millimetres or less per day) governing the initiation of recharge. These thresholds tend to rise as aridity increases, and recharge in drylands is more episodic and increasingly dominated by focused recharge through losses from ephemeral overland flows. Extreme annual recharge is commonly associated with intense rainfall and flooding events, themselves often driven by large-scale climate controls. Intense precipitation, even during years of lower overall precipitation, produces some of the largest years of recharge in some dry subtropical locations. Our results therefore challenge the 'high certainty' consensus regarding decreasing water resources
in such regions of sub-Saharan Africa. The potential resilience of groundwater to climate variability in many areas that is revealed by these precipitation-recharge relationships is essential for informing reliable predictions of climate-change impacts and adaptation strategies.
Groundwater ecosystems remain poorly understood yet may provide ecosystem services, make a unique contribution to biodiversity and contain useful bio-indicators of water quality. Little is known ...about ecosystem variability, the distribution of invertebrates within aquifers, or how representative boreholes are of aquifers. We addressed these issues using borehole imaging and single borehole dilution tests to identify three potential aquifer habitats (fractures, fissures or conduits) intercepted by two Chalk boreholes at different depths beneath the surface (34 to 98 m). These habitats were characterised by sampling the invertebrates, microbiology and hydrochemistry using a packer system to isolate them. Samples were taken with progressively increasing pumped volume to assess differences between borehole and aquifer communities. The study provides a new conceptual framework to infer the origin of water, invertebrates and microbes sampled from boreholes. It demonstrates that pumping 5 m(3) at 0.4-1.8 l/sec was sufficient to entrain invertebrates from five to tens of metres into the aquifer during these packer tests. Invertebrates and bacteria were more abundant in the boreholes than in the aquifer, with associated water chemistry variations indicating that boreholes act as sites of enhanced biogeochemical cycling. There was some variability in invertebrate abundance and bacterial community structure between habitats, indicating ecological heterogeneity within the aquifer. However, invertebrates were captured in all aquifer samples, and bacterial abundance, major ion chemistry and dissolved oxygen remained similar. Therefore the study demonstrates that in the Chalk, ecosystems comprising bacteria and invertebrates extend from around the water table to 70 m below it. Hydrogeological techniques provide excellent scope for tackling outstanding questions in groundwater ecology, provided an appropriate conceptual hydrogeological understanding is applied.
We assess the use of fluorescent dissolved organic matter at excitation-emission wavelengths of 280nm and 360nm, termed tryptophan-like fluorescence (TLF), as an indicator of faecally contaminated ...drinking water. A significant logistic regression model was developed using TLF as a predictor of thermotolerant coliforms (TTCs) using data from groundwater- and surface water-derived drinking water sources in India, Malawi, South Africa and Zambia. A TLF threshold of 1.3ppb dissolved tryptophan was selected to classify TTC contamination. Validation of the TLF threshold indicated a false-negative error rate of 15% and a false-positive error rate of 18%. The threshold was unsuccessful at classifying contaminated sources containing <10 TTC cfu per 100mL, which we consider the current limit of detection. If only sources above this limit were classified, the false-negative error rate was very low at 4%. TLF intensity was very strongly correlated with TTC concentration (ρs=0.80). A higher threshold of 6.9ppb dissolved tryptophan is proposed to indicate heavily contaminated sources (≥100 TTC cfu per 100mL). Current commercially available fluorimeters are easy-to-use, suitable for use online and in remote environments, require neither reagents nor consumables, and crucially provide an instantaneous reading. TLF measurements are not appreciably impaired by common intereferents, such as pH, turbidity and temperature, within typical natural ranges. The technology is a viable option for the real-time screening of faecally contaminated drinking water globally.
Display omitted
•Urgent need to screen drinking water for faecal contamination rapidly.•Multi-country assessment of TLF as an indicator of faecal contamination.•A 1.3ppb dissolved tryptophan threshold is effective to infer contamination.•TLF is strongly correlated with thermotolerant coliform concentration.•TLF is a commercially available, easy-to-use, reagentless, real-time option.
•TLF/HLF more related to thermotolerant coliforms (TTCs) than other rapid approaches.•Relationships between TLF/HLF and TTCs are significant in the wet season.•TLF is significantly related to total ...bacterial counts, not TTCs, in the dry season.•TLF/HLF at any point in time relate to wet season TTCs, when TTCs are elevated.•In-situ TLF/HLF are more resilient faecal contamination risk indicators than TTCs.
Faecal indicator organisms (FIOs) are limited in their ability to protect public health from the microbial contamination of drinking water because of their transience and time required to deliver a result. We evaluated alternative rapid, and potentially more resilient, approaches against a benchmark FIO of thermotolerant coliforms (TTCs) to characterise faecal contamination over 14 months at 40 groundwater sources in a Ugandan town. Rapid approaches included: in-situ tryptophan-like fluorescence (TLF), humic-like fluorescence (HLF), turbidity; sanitary inspections; and total bacterial cells by flow cytometry. TTCs varied widely in six sampling visits: a third of sources tested both positive and negative, 50% of sources had a range of at least 720 cfu/100 mL, and a two-day heavy rainfall event increased median TTCs five-fold. Using source medians, TLF was the best predictor in logistic regression models of TTCs ≥10 cfu/100 mL (AUC 0.88) and best correlated to TTC enumeration (ρs 0.81), with HLF performing similarly. Relationships between TLF or HLF and TTCs were stronger in the wet season than the dry season, when TLF and HLF were instead more associated with total bacterial cells. Source rank-order between sampling rounds was considerably more consistent, according to cross-correlations, using TLF or HLF (min ρs 0.81) than TTCs (min ρs 0.34). Furthermore, dry season TLF and HLF cross-correlated more strongly (ρs 0.68) than dry season TTCs (ρs 0.50) with wet season TTCs, when TTCs were elevated. In-situ TLF or HLF are more rapid and resilient indicators of faecal contamination risk than TTCs.
•Most common viruses were hepatitis A virus, norovirus GI and hepatitis E virus.•Viruses most prevalent and at higher concentrations during the main recharge season.•89% of positive detections ...occurred during a rising water table.•Seasonality of norovirus GI detections matches seasonal incidence in population.•No evidence for use of bacterial or viral indicator organisms at the sample level.
We investigated the seasonal prevalence of seven enteric viruses in groundwater-derived public water sources distributed across the dominant aquifers of England. Sampling targeted four periods in the hydrological cycle with typically varying microbial risks, as indicated using a decade of Escherichia coli prevalence data. Viruses were concentrated onsite by filtration of raw groundwater, and extracted nucleic acid (NA) was amplified by qPCR or RT-qPCR. Seven out of eight sources, all aquifers, and 31% of samples were positive for viral NA. The most frequently detected viral NA targets were Hepatitis A virus (17% samples, 63% sites), norovirus GI (14% samples, 38% sites), and Hepatitis E virus (7% samples, 25% sites). Viral NA presence was episodic, being most prevalent and at its highest concentration during November and January, the main groundwater recharge season, with 89% of all positive detects occurring during a rising water table. Seasonal norovirus NA detections matched its seasonal incidence within the population. Viral NA is arriving with groundwater recharge, as opposed to persisting for long-periods within the saturated zone. Neither total coliforms nor E. coli were significant predictors of viral NA presence-absence, and there was limited co-occurrence between viruses. Nevertheless, a source with an absence of E. coli in regularly collected historical data is unlikely to be at risk of viral contamination. To manage potential groundwater viral contamination via risk assessment, larger scale studies are required to understand key risk factors, with the evidence here suggesting viral NA is widespread across a range of typical microbial risk settings.
Fluorescent natural organic matter at tryptophan-like (TLF) and humic-like fluorescence (HLF) peaks is associated with the presence and enumeration of faecal indicator bacteria in groundwater. We ...hypothesise, however, that it is predominantly extracellular material that fluoresces at these wavelengths, not bacterial cells. We quantified total (unfiltered) and extracellular (filtered at < 0.22 µm) TLF and HLF in 140 groundwater sources across a range of urban population densities in Kenya, Malawi, Senegal, and Uganda. Where changes in fluorescence occurred following filtration they were correlated with potential controlling variables. A significant reduction in TLF following filtration (ΔTLF) was observed across the entire dataset, although the majority of the signal remained and thus considered extracellular (median 96.9%). ΔTLF was only significant in more urbanised study areas where TLF was greatest. Beneath Dakar, Senegal, ΔTLF was significantly correlated to total bacterial cells (ρ
0.51). No significant change in HLF following filtration across all data indicates these fluorophores are extracellular. Our results suggest that TLF and HLF are more mobile than faecal indicator bacteria and larger pathogens in groundwater, as the predominantly extracellular fluorophores are less prone to straining. Consequently, TLF/HLF are more precautionary indicators of microbial risks than faecal indicator bacteria in groundwater-derived drinking water.
We explore in-situ fluorescence spectroscopy as an instantaneous indicator of total bacterial abundance and faecal contamination in drinking water. Eighty-four samples were collected outside of the ...recharge season from groundwater-derived water sources in Dakar, Senegal. Samples were analysed for tryptophan-like (TLF) and humic-like (HLF) fluorescence in-situ, total bacterial cells by flow cytometry, and potential indicators of faecal contamination such as thermotolerant coliforms (TTCs), nitrate, and in a subset of 22 samples, dissolved organic carbon (DOC). Significant single-predictor linear regression models demonstrated that total bacterial cells were the most effective predictor of TLF, followed by on-site sanitation density; TTCs were not a significant predictor. An optimum multiple-predictor model of TLF incorporated total bacterial cells, nitrate, nitrite, on-site sanitation density, and sulphate (r2 0.68). HLF was similarly related to the same parameters as TLF, with total bacterial cells being the best correlated (ρs 0.64). In the subset of 22 sources, DOC clustered with TLF, HLF, and total bacterial cells, and a linear regression model demonstrated HLF was the best predictor of DOC (r2 0.84). The intergranular nature of the aquifer, timing of the study, and/or non-uniqueness of the signal to TTCs can explain the significant associations between TLF/HLF and indicators of faecal contamination such as on-site sanitation density and nutrients but not TTCs. The bacterial population that relates to TLF/HLF is likely to be a subsurface community that develops in-situ based on the availability of organic matter originating from faecal sources. In-situ fluorescence spectroscopy instantly indicates a drinking water source is impacted by faecal contamination but it remains unclear how that relates specifically to microbial risk in this setting.
Display omitted
•Total bacterial cells most related variable to tryptophan-like fluorescence (TLF)•TLF and humic-like fluorescence strongly correlate with dissolved organic carbon.•Thermotolerant coliforms are not strongly related to other variables.•TLF and HLF relate to faecal contamination.
The impact of climate variability on groundwater storage has received
limited attention despite widespread dependence on groundwater as a resource
for drinking water, agriculture and industry. Here, ...we assess the climate
anomalies that occurred over Southern Africa (SA) and East Africa, south of
the Equator (EASE), during the major El Niño event of 2015–2016, and their
associated impacts on groundwater storage, across scales, through analysis
of in situ groundwater piezometry and Gravity Recovery and Climate Experiment (GRACE)
satellite data. At the continental scale, the El Niño of 2015–2016 was associated with a
pronounced dipole of opposing rainfall anomalies over EASE and Southern
Africa, north–south of ∼12∘ S, a characteristic pattern
of the El Niño–Southern Oscillation (ENSO). Over Southern Africa the most intense drought event in the
historical record occurred, based on an analysis of the cross-scale areal
intensity of surface water balance anomalies (as represented by the
standardised precipitation evapotranspiration index – SPEI), with an
estimated return period of at least 200 years and a best estimate of
260 years. Climate risks are changing, and we estimate that anthropogenic warming
only (ignoring changes to other climate variables, e.g. precipitation) has
approximately doubled the risk of such an extreme SPEI drought event. These
surface water balance deficits suppressed groundwater recharge, leading to a
substantial groundwater storage decline indicated by both GRACE satellite
and piezometric data in the Limpopo basin. Conversely, over EASE during the
2015–2016 El Niño event, anomalously wet conditions were observed with an
estimated return period of ∼10 years, likely moderated by the
absence of a strongly positive Indian Ocean zonal mode phase. The strong but
not extreme rainy season increased groundwater storage, as shown by satellite
GRACE data and rising groundwater levels observed at a site in central
Tanzania. We note substantial uncertainties in separating groundwater from
total water storage in GRACE data and show that consistency between GRACE
and piezometric estimates of groundwater storage is apparent when spatial
averaging scales are comparable. These results have implications for
sustainable and climate-resilient groundwater resource management, including
the potential for adaptive strategies, such as managed aquifer recharge
during episodic recharge events.
•Modeling sensitivity analysis performed for 4 soil moisture content monitoring sites.•Soil texture uncertainty leads to significant groundwater recharge uncertainty.•Aggregated soil moisture content ...is virtually insensitive to soil texture.
Estimating groundwater recharge rates is important for water resource management studies. Modeling approaches to forecast groundwater recharge typically require observed historic data to assist calibration. It is generally not possible to observe groundwater recharge rates directly. Therefore, in the past, much effort has been invested to record soil moisture content (SMC) data, which can be used in a water balance calculation to estimate groundwater recharge. In this context, SMC data is measured at different depths and then typically integrated with respect to depth to obtain a single set of aggregated SMC values, which are used as an estimate of the total water stored within a given soil profile. This article seeks to investigate the value of such aggregated SMC data for conditioning groundwater recharge models in this respect. A simple modeling approach is adopted, which utilizes an emulation of Richards’ equation in conjunction with a soil texture pedotransfer function. The only unknown parameters are soil texture. Monte Carlo simulation is performed for four different SMC monitoring sites. The model is used to estimate both aggregated SMC and groundwater recharge. The impact of conditioning the model to the aggregated SMC data is then explored in terms of its ability to reduce the uncertainty associated with recharge estimation. Whilst uncertainty in soil texture can lead to significant uncertainty in groundwater recharge estimation, it is found that aggregated SMC is virtually insensitive to soil texture.