Thank you for visiting Nature. The browser version you are using has limited support for CSS. For the best experience, we recommend that you use a newer version of the browser (or turn off the compatibility mode in Internet Explorer). At the same time, to ensure continued support, we will display sites without styles and JavaScript.

Nature Volume 598, Pages 611–617 (2021) Cite this article

Access to safely managed drinking water (SMDW) remains a global challenge, affecting 2.2 billion people1,2. Solar-powered atmospheric water harvesting (AWH) equipment with continuous cycles can accelerate progress by extracting water from the air by dispersing 3, 4, 5, 6, but low specific yield (SY) and low daytime relative humidity (RH) are proposed The following questions their performance (in terms of daily water output) 7, 8, 9, 10, 11. However, as far as we know, despite the superior conditions in the tropics, there is no analysis to delineate the global potential of AWH12, where two-thirds of people do not have SMDW2. Here, we show that AWH can provide SMDW to 1 billion people. Our evaluation (using Google Earth Engine13) introduced a hypothetical 1 square meter device at 30% to 90% RH, respectively. Such equipment can meet the target average daily drinking water demand of 5 liters per person per day14. We map the impact potential of existing equipment and new adsorbent categories, which shows that these goals can be achieved through continuous technological development and are within thermodynamic limits. In fact, these performance goals have been achieved through experiments in the demonstration of adsorbent materials 15, 16, 17. Our tools can provide information on the design and trade-offs of atmospheric water collection equipment to maximize global impact, while constantly striving to use existing technologies to achieve the Sustainable Development Goals (SDG).

Ensuring that everyone has reliable access to safe drinking water remains a global challenge and is officially recognized as an international development priority by 2030 in the United Nations Global Development Priorities Framework Sustainable Development Goal 6.118. Progress towards this goal is measured by the WHO/UNICEF Joint Monitoring Program (JMP) as the percentage of the population using Safely Managed Drinking Water (SMDW), where “safety management” is defined as “improved water sources located within the premises” , When needed and without feces and priority chemical pollution"1,2. The traditional route of bringing SMDW to the currently unserved population is estimated to cost US$114 billion a year (since 2015), which is more than three times the historical financing trend19. In addition, there is a growing global interest in solutions that provide safe drinking water without the environmental impact of increasing reliance on bottled water, and does not require family-level intervention, which has limited compliance20,21 . If cost-effective off-grid equipment can be designed and expanded, atmospheric water harvesting (AWH) is expected to accelerate decentralized access to underserved communities.

Several types of off-grid AWH designs exist or are being explored 8, 12, 22, and 23, as summarized in Table 1. AWH equipment is classified by energy-active equipment uses external energy, while passive equipment only relies on being allowed to collect pre-condensed dew or mist. Therefore, passive devices are limited to geographic environments where dew or mist can be collected systematically7,12,24. Active adsorbent-based AWH equipment mainly uses solar thermal energy to extract water in one of two modes of operation: day and night mode equipment extracts at night (when relative humidity is high) and during daytime condensation (when solar energy is available) once a day, one cycle is required Large adsorbent bed. In contrast, continuous mode equipment is not limited to a single daily cycle, only a small amount of water vapor needs to be maintained during the process3,4, which greatly reduces the quality of the adsorbent and the size of the equipment. However, this requires extraction with a lower relative humidity when solar energy is available, which can cause performance issues7,8,9,10,11. Cooler-condenser equipment uses power (usually electricity) to actively cool the air below the dew point and collect condensate. If it is driven by solar energy, photovoltaic (PV) panels are required. Unlike solar-thermal energy equipment, solar-powered cooler-condenser equipment has a sharp loss in electrical energy conversion. In the context of a specific output, we use kWh to represent the primary solar energy before heat loss and other losses, and kWhPV to represent the electrical energy provided to the device from the PV panel after conversion. Unless otherwise stated, the range of SY refers to the RH between 30% and 90% at 20 °C.

Here, we use global data to evaluate the solar-driven continuous mode AWH (SC-AWH). The SY of AWH is much lower than the water source of infrastructure such as desalination (about 200 l kWh-1). However, the SC-AWH equipment with a size that can provide enough daily drinking water output for individuals or families can solve the water quality and water supply dimensions of the household-level SMDW solution.

In order to estimate the impact potential of SC-AWH, we first mapped the distribution of approximately 2.2 billion people without SMDW2. Recent studies have used geostatistical techniques to estimate inequalities in local safe drinking water and sanitation facilities from various data sources26,27 reporting facility type indicators. Here, we used a deterministic method based entirely on JMP drinking water service level data. In this study, we assume that SC-AWH is only used for drinking water and cannot replace water for other household uses, such as sanitation, cooking, and sanitation14,28.

Figure 1a shows the overall percentage of the population in the areas with the lowest levels of each available area reported by JMP. This seamless structure of national and sub-national survey areas provides a spatial continuous map of the global population distribution without SMDW. Consistent with previous reports2,29, sub-Saharan Africa has the highest total population without SMDW, followed by South Asia and Latin America.

a. The percentage of the population without SMDW in the survey area reported by the WHO/UNICEF JMP. b. Record the population density of people without SMDW from WorldPop at a resolution of 1 km, adjusted by the JMP ratio at a resolution of 1 km. Generated in ArcGIS 10.

The area ratio in Figure 1a is applied as a linear weight to each pixel of the WorldPop (2017) 1km resolution residential population count image (https://www.worldpop.org). This provides an estimate of the spatial resolution of the distribution of people without SMDW, which is closer to the scale at which climate variables related to AWH change due to physical geography (such as topography and land cover). The resulting weighted population distribution is shown in Figure 1b.

We propose a geospatial tool (AWH-Geo) to assess the global potential of the nominal SC-AWH device given the available climate resources. AWH-Geo is built in Google Earth Engine13 and can be extended across climate data. In this study, AWH-Geo used the ERA5-Land climate reanalysis during the 10-year period 2010-2019 (inclusive). ERA5-Land was selected for its high resolution (9 kilometers per hour interval), global coverage, and ability to represent historical weather conditions. Although the inter-decadal trends are briefly discussed in Figure 9 of the extended data, the cycle is sufficient to explain the inter-annual changes. In order to run your own analysis in a shorter calculation time, users can adjust the analysis cycle in the tool.

AWH-Geo takes the instantaneous rate of water output as a function of three main environmental variables as input: (1) global horizontal irradiance from sunlight (GHI (W m-2)), (2) RH (%) and (3) Air temperature (T(°C)). Secondary climate variables can be included later (for example, downstream infrared and surface wind speed). We propose an output table in which the water production value is a function of the bin climate inputs GHI, RH, and T as a way to link the AWH equipment model or experimental characteristics with the geospatial analysis. The water output can be entered as an area collection rate (in l h-1 m-2) for water withdrawal, or as the expected output (in l h-1) of a real device with a known collection area. Among all the data points of the multi-year climate image time series, AWH-Geo uses the given output table to find production values ​​and aggregates the water output for display as a global map or derived map. Although previous assessments were limited to relatively few locations with on-site meteorological data7,30 or to an area analysis31, the method presented here is global and spatially continuous. Figure 2 shows the conceptual workflow of the AWH-Geo and adjacent processes that produced the results in this study.

The cylinder represents data storage from Google Earth Engine, WHO/UNICEF JMP or open online content. Shown are the process (rectangle), geographic image (parallelogram), and output (circle).

We first use AWH-Geo to plot the theoretical upper limit of solar-powered AWH by constructing the output table in the literature as a specific water production SY (in l kWh-1). SY is an assessment index of AWH's sensitivity to RH32 and is the inverse of specific energy consumption (SEC), which is often used in other water and seawater desalination systems. The resulting map overlaps with the point density of the distribution of people without SMDW, which is used for the visual comparison in Figure 3.

ac, the average daily water output of solar-powered AWH, given the overall thermodynamic limit of any process (Thot = 100 °C) (a), the chiller-condenser process driven by PV32 (b) and the active adsorbent device Examples of types (TRP) are from the gel in Reference 15) (c). The annotated chart in the illustration shows the selection seasonal profile with average output and bi-weekly intervals of the main climate drivers: GHI, RH, and temperature. The output (in l d-1 m-2) is normalized to the horizontal device area under sunlight. The performance of the actual equipment will be lower than the maximum theoretical potential. The superimposed dot density of 2.2 billion people without SMDW (red). The placement of the points in the entire survey area is spatially arbitrary. Generated in ArcGIS 10.

Recently, Jin et al. The basic thermodynamic limitations of AWH33 have been described. The model gives the minimum thermal energy required per unit water output of the black box AWH (at a given hot side temperature level), corresponding to a value of SY between 5 and 50 l kWh-1. Kim's thermodynamic limit is shown in Figure 3a. Plotting the thermodynamic limits helps to set the maximum expectations for SC-AWH output worldwide and to assess the potential for improvement that may exist between the performance of existing equipment and the basic physical limits. Similar analytical methods have been used to evaluate condenser-based equipment, day and night equipment, and dew collectors for specific locations or areas7, 12, 30, 31. The output geographic pattern usually closely follows the time average humidity value and is affected by sunlight. It is worth noting that the results show that most parts of the world have huge water production potential, especially in tropical regions.

Next, we mapped the maximum output of the two basic design types. Peeters describes the maximum output of the active cooler-condenser, giving SY 1-30 \({\rm{l}}\,{{\rm{kWh}}}_{{\rm{PV}}} ^{- 1}\) (0.2–6 l kWh−1), drawn using AWH-Geo in Figure 3b. For the adsorbent design, metal organic framework (MOF) and thermally responsive polymer (TRP) gels showed the highest yields at low and high relative humidity, respectively. Zhao et al. It proves the excellent performance of TRP15 at high RH (0.2-9.3 l kWh-1 (converted from Peeters32 to SY)), which is generally better than MOF (its reported maximum 32 SY is approximately 1 l kWh-1). Zhao's TRP global forecast is mapped in Figure 3c.

In addition to the annual average, AWH-Geo can also derive indicators useful for analyzing seasonal changes in output. Or, AWH-Geo derives a 90% availability (P90) value in a set of time windows (methods).

Our coincidence analysis calculated the average number of hours per day when GHI and RH were both above the parameter threshold. Figure 4a plots the annual average of such daily coincident hours for a given threshold pair, interpreted as the daily operating hours (ophd) of the hypothetical equipment. The important transition zone between tropical and desert regions shows the expected balance between sunlight and humidity, which is usually inversely proportional. In the arid Sahel, a very low RH threshold of 10% increases the ophd potential by only 1-2 hours from the ophd of 30% RH, exceeding the GHI threshold, but the ophd then drops sharply at the higher RH threshold. This indicates diminishing returns for equipment operating below 30%. Coastal areas are expected to maintain a consistent 2-4 ophd above 50% RH in the world.

a, b, the geographical distribution (a) and sum (b) of the non-SMDW population living in areas that meet the parameter thresholds related to the operation of SC-AWH equipment. The number of operating hours per day (Ophd) is the average daily duration that exceeds both the sunshine (GHI) and RH thresholds. Example of use: A device that requires more than 5 h d-1 of sunlight over 400 W m-2 must operate at as low as 40% RH to cover approximately 700 million users. c, d, the average daily output achievable by people without SMDW is normalized to the horizontal equipment area under sunlight (c) and the SY profile (d). The target curve is a hypothetical SY profile that can provide 5 l d-1 for a given solar energy collection area. The water output and the SY target are linearly proportional to the area of ​​the equipment in the sun. Therefore, for demonstration purposes, we show that for a given RH, doubling the area of ​​the equipment from 1 square meter to 2 square meters will halve the SMDW target SY requirement for the target population. The ZMW source configuration file approximates the manufacturer's technical specification sheet 35. Please note that the complete ZMW panel is approximately 3 square meters. The experimental values ​​of MOF and adsorbent are taken from experiment 3,36 (0.19 l kWh-1 and 0.84 l kWh-1), and TRP is taken from references. 15. All converted to reference. 32. The value of Bagheri device 34 assumes work rather than heat input; therefore, photovoltaic efficiency is applied when converting from GHI. The map was generated in ArcGIS 10.

Next, we use the weighted population image to sum the populations who have not visited the SMDW, and accumulate the groups by ophd at the entire interval, as shown in Figure 4b. When the RH value is between 30% and 50%, the GHI value is between 400 and 600 W m-2, and the ophd value is between 3 and 5 hours, the user potential gradually decreases. These reflect the key temporal and spatial population patterns of similar climate transitions in the tropics, where most people without SMDW live-especially in the savannas of sub-Saharan Africa and the Ganges river basin in India. Devices that can operate above these values ​​theoretically have the potential to serve more than half of the world's remaining population who cannot obtain SMDW.

Next, we ran a series of SY curves of SY curves through AWH-Geo, including data sheets for commercial chiller-condenser equipment and SOURCE panels evaluated by Bagheri34, SOURCE’s adsorbent-based equipment, formerly known as zero Quality water 35 (ZMW).

Figure 4c shows the output normalized by area (in l d-1 m-2)-the performance index advocated by LaPotin et al. 11-As a function of population not reaching SMDW. The steep gradient of the output's impact on humans reflects the gradient in the coincidence analysis. The linear SY profile prioritizes performance under low RH, but can limit output even in resource-rich climates. The target curve is based on a hypothetical SY value, similar to the characteristics of an adsorbent or device profile, with an average of 5 l d-1 m-2 reaching 1 billion users. Comparing the two target curves shows the expected trade-off between serving more users with low output (linear) and serving fewer users with high output (logical).

In order to further explore the trade-off between different RH values ​​of the SY curve, we have drawn the relationship between the SY value from materials and equipment and the target curve, so as to reach 50-2 million people at 5 ld-1 without SMD-1, that is The amount of alcohol consumed per day is approximately the individual's water demand (Figure 4d). Unless otherwise stated, we base the target curve on a 1 square meter device, although the water output and SY target are linearly proportional to the device area in the sun. To prove this, we drew a version of a 1 billion target based on 2 m2-doubling the area of ​​the equipment halved the SY requirement for the target's impact. Existing equipment all follow an approximately linear output under the RH target curve of less than 500 million shocks. MOF and other adsorbents show different results3,36, although they remain roughly linear. Zhao's outstanding earnings under high RH compensated for the low performance under low RH (logistics profile), and showed the most hope to reach the largest user base (2 billion). Figure 4d compares the material and equipment performance side by side to show the gap between the current capability and the theoretical limit, although the actual equipment will suffer losses and thus cannot fully reach the ideal material performance or theoretical limit.

This study came up with preliminary conclusions-the development of detailed SC-AWH design standards requires further work. A device with a solar collection area of ​​1 square meter and a SY profile of 0.2-2.5 l kWh-1 (2 square meters is 0.1-1.25 l kWh-1) can meet the SMDW needs of about 1 billion people, assuming continuous collection 2–daily 3 The hourly coincident sunlight exceeds 600 W m−2 and the relative humidity is higher than 30%. The shape of the SY curve is essential for SC-AWH to utilize consistent humidity and solar energy during key periods of the day (usually morning and evening). There is a trade-off between increasing production in the climate transition zone (northern sub-Saharan Africa and western India) with low relative humidity (about 30%) and focusing on the exponential increase in production in humid regions such as Bangladesh and the equatorial region.

Researchers and equipment inventors can cross-refer to Figure 4 when making trade-off decisions between technical specification sets and serviceable areas and personnel. Recent experiments 4,5,37 show the rapid increase in the output of multi-cycle adsorbents, from 0.1 to more than 8.0 ld−1 kg−1 under outdoor conditions (RH 10–60%, GHI <1,000 W m-2). , And show performance changes in a range similar to the population distribution 11,31 (RH 30-50%, GHI 400-600 W m-2). The equipment efficiency improvement from the innovative design architecture 38 and the new high-performance physical adsorbent 15, 17, 39, 40, 41 shows the prospect of increasing the output of SC-AWH. Individual specific yields from material experiments or prototypes can be plotted in Figure 4d for benchmarking against target impact. Equipment performance and published output tables verified under outdoor field conditions are necessary for researchers around the world to advance the progress of AWH.

The long-term average output of AWH equipment is an important but limited indicator. Seasonal, weekly and day-night changes in production will affect user adoption and market viability. Some seasonal profiles are explored in the expanded data graph. 4-8. Short-term shortages can be replenished by previous surplus storage. During periods of seasonal shortages, such as monsoon climates, rainfall or alternative sources need to be collected. The use of multiple water sources and seasonal shifts have been well documented in the literature, although there may be trade-offs in water quality and pollution42,43, which reinforces the need for in-depth understanding of existing water extraction practices when deploying AWH, and Focus on household water treatment and safe storage.

Considering the scale of the global atmospheric water budget, the hydro-ecological impact of AWH on drinking water may be negligible. The total amount of services provided to all 2.2 billion people without SMDW for 10 ld-1 is approximately 8 km3 yr-1, which is only 0.20% of global farmland water extraction (4,000 km3 yr-1) and total land evaporation 0.01D (65,500 km3 yr−1).

SC-AWH equipment has low cost potential. Most design architectures have few moving parts (for example, the slowly rotating adsorbent wheel 8), and can be composed of widely available components. Advanced adsorbent materials (for example, MOF or TRP) need to be manufactured on a large scale to meet cost targets. The new high-volume manufacturing method of MOF45,46 has the potential to significantly reduce costs.

Technological development is only one part of the complex issue of safe water extraction; user-centered formative research and a wide range of end users are essential to ensure that equipment is widely adopted. A device similar to bottled water 21 SC_AWH may paradoxically undermine efforts to develop permanent pipeline infrastructure. Product affordability and adoption require parallel financial and socio-cultural efforts, such as expanding loan availability, raising awareness of the risks of water-borne diseases, and increasing women’s influence on community decision-making 47,48,49.

Our analysis shows that daytime climatic conditions may actually be sufficient to meet the continuous mode AWH operation in the regions of the world where human needs are highest. The evaluation shows that it is worthwhile to focus equipment design standards on the maximum impact and cost reduction of household-scale off-grid drinking water production.

Drinking water coverage data by region comes from WHO/UNICEF JMP. JMP is the official custodian of global data on water supply, sanitation and personal hygiene2. It absorbs administrative data from various countries, national census and survey data, and maintains a database that can be accessed online through its website. We accessed data tables on national and local drinking water service levels from https://washdata.org.

The JMP dataset is not geographically related to official boundary documents. We added the table to the GIS boundaries obtained from the following open source collections: GADM (https://gadm.org), the spatial data repository of the U.S. Agency for International Development (DHS) Demographic and Health Survey Program, and Radboud Global Data Laboratory University (GDL) 2,50,51,52,53. The sub-national regions reported by JMP are unstructured and represent different regional administrative levels (province, state, district, etc.).

Use custom geoprocessing tools built into Python and ArcGIS 10 to add JMP country and subnational data to the GIS boundary. The tool adds the available JMP sub-national survey data to the closest name match (admin1, admin2, and admin3) of the regional boundary names from the GADM merged stack, global DHS and GDL boundaries. For countries where no local data is available, the JMP national survey data is then added to the GADM country (admin0) border. Finally, the two border-connected datasets (country and sub-country) are merged, processed, and exported into a seamless global water stress population data structure with the highest spatial resolution available to each (Figure 1a).

JMP does not report the breakdown between SMDW and basic service levels in subnational areas, but instead reports a combined category called "at least basic" (ALB). In order to estimate the SMDW value of the sub-national region, a simple cross-multiplication was performed using the country-level segmentation:

Among them, ALBnational, ALBsubnational, and SMDWnational are known values.

The verification of the cross-estimation of ALB's share of SMDW in sub-countries and regions was carried out on the reference data set of the national representative household survey, which collected all the standard data of SMDW54, as shown in Figure 2 of the extended data. We reported R2 = 0.87 and standard errors of 3.67, indicating that there is a bias in our study that overreports the share of SMDW and may underestimate people without SMDW. This difference comes from SMDW's JMP calculation, which relies on the minimum value of multiple drinking water service standards (no pollution, available when needed, and site accessible), rather than considering whether a single household meets all the standards of SMDW55.

The proportion of the population without SMDW is multiplied by the resident population value of 1 km spatial resolution 56 (https://www.worldpop.org) in the 2017 WorldPop top-down unconstrained global mosaic population count. WorldPop is accessed online as a TIF image and imported into Google Earth Engine. 2017 was chosen to more closely match JMP's water withdrawal data. The percentage reported by JMP may not be uniform in most regions. 57 brings unknown error to Figure 1b, but given the limitations of the reported data in these regions, it represents the best estimate available to us.

We use GHI (in W m−2) as the solar input data. GHI has good usability in climate data sets and introduces the fewest number of assumptions. Since GHI describes the irradiance in a local horizontal reference plane, this approximation is only applicable to devices with a horizontal solar collection area. The annual average comparison between horizontal and best fixed tilt panels shows that the difference between direct and diffuse radiation at tropical latitudes is negligible, while at locations within 50° north-south latitude, the ratio is less than 25X. Those seeking accurate absolute predictions for tilting equipment or higher latitudes are encouraged to adjust the provided code according to their specific assumptions.

As discussed in the main text, solar-powered AWH devices usually have one of two main energy inputs: heat (converted directly from incident sunlight on the device) or electricity (from PV). Here, the energy unit used to calculate the output in l kWh-1 is the incident solar energy directly from GHI. Various assumptions are related to the reported value based on their source. Assume that the thermal limit 33, target curve and experimental results reported by TRP15 and MOF have the ability to convert sunlight directly (100%) into heat. For the ZMW device, the table provided by the manufacturer takes into account the system loss, so the table values ​​are directly converted in our model. For reference. 34 and the cooler-condenser limit in the references. In Figure 32, they all assume input power instead of heat. We applied a typical PV conversion efficiency of 20% to convert solar kWh (GHI) into kWh PV (electrical power) and input to the device 59.

AWH-Geo assumes continuous or quasi-continuous AWH. AWH-Geo considers each 1-hour time step independently, so it is stateless. Except for edge cases, this is a safe assumption for high-quality SC-AWH equipment. For adsorbent cycles and most thermal time constants, the time constants of these equipment are usually less than 1 hour. For equipment with long time constants, batch equipment, or processes with slow (de)sorption kinetics, this assumption may lead to increased errors and may require further adjustments to the provided code.

AWH-Geo is AWH's resource assessment tool. It consists of a geospatial processing pipeline, which is used to map the water production (in liters per unit time) of any solar-powered continuous AWH equipment around the world. Its characteristic is that the output table is in the form of output = f(RH, T , GHI).

The output table shows the AWH output values ​​of the 3 main climate variables arranged in the following range, in units of lh−1 or lh−1 m−2: RH is between 0 and 100%, with an interval of 10%, and GHI is between 0 Between 2 and 1,300 W m−2 at 100 W m-2 intervals, T between 0 and 45 °C, at 2.5 °C intervals (2,145 total output values). These tables are converted into 3D array images in Google Earth Engine and processed in a collection of climate time series images of the period of interest. Finally, these AWH output values ​​are synthesized (reduced) into a single time average statistic of interest as an image.

The climate time series data comes from the ERA5-terrestrial climate reanalysis of the European Centre for Medium-Range Weather Forecast (ECMWF) 60, which can be accessed from the Google Earth Engine data directory. ERA5-Surface variables are used at 1 hour intervals and 0.1°×0.1° (nominal 9 kilometers). This work used a 10-year analysis period (2010-2019, including 2010-2019), which represents a time period sufficient to provide a reasonable correction for the medium-term interannual climate variability.

The climate variables GHI and T match the ERA5-Land parameters "surface solar radiation down" (converted from cumulative to hourly average) and "2m temperature" (converted from K to °C), respectively. RH is calculated based on the ambient temperature and dew point temperature parameters. Its relationship is derived from August-Roche-Magnus approximately 61, rearranged as:

Where a is 17.625 (constant), b is 243.04 (constant), T is the ERA5-Land parameter "2m temperature" converted from K to °C, and Td is from K to °C.

In 2016, in Ames, Iowa (using the Iowa environmental mediation network AMES-8-WSW station 62), manually performed field verification of multiple time steps of climate parameters and mapping output in Google Earth Engine , And showed trivial errors (< 5%).

Figure 3a plots the thermodynamic upper limit output of SC-AWH based on the equation of Kim et al. 33, reproduced below.

Where \({\dot{Q}}_{{\rm{hot}},{\rm{in}},{\rm{\min }}}\) is the required minimum input heat flux (in Wheat as a unit) drives the process, \({T}_{{\rm{hot}}}\) is the temperature (in K) when the input heat is transferred, \({T}_{{\rm{ ambient} }}\) is the ambient temperature (in K), at which temperature, heat is discharged and water and air leave the process, \({\dot{m}}_{{\rm{water}},{ \rm{out} }}\) is the mass production rate of liquid water, \(\omega \) represents the humidity ratio of kilograms of water per kilogram of dry air, \({e}\) represents the specific energy, you can find Set the temperature and humidity, the subscript air,in means that the ambient air inhaled at \({T}_{{\rm{ambient}}}\) extracts moisture from it, and the subscript air,out means that it is in \( {T} _{{\rm{ambient}}}\) After extracting some water from it, the subscript water,out means that the liquid water is in \({T}_{{\rm{ambient}}}\) as the required product.

Parameters that do not exist in this formula, but are in Kim's basic derivation: This upper limit is obtained with a small recovery rate (RR ~ 0) selected for numerical stability and reversible process conditions (entropy generation, Sgen = 0).

Kim's model assumes AWH, where the basic energy required is driven by the input heat provided at temperature \({T}_{{\rm{hot}}}\). As long as the heat drives the process, the limitation it represents is independent of the process, the number of stages, the choice of adsorbent, etc.

We apply Kim's model to solar input, assuming that the ideal conversion efficiency from solar irradiance to available heat is 100%. This idealization retains a robust upper limit without introducing additional parameters. For heat sinks, the literature value of the theoretical solar thermal efficiency limit ranges from >99.99 to 95.80%, depending on the level of angular selectivity63.

Rearranged, Kim’s model is generated

Among them, \({\dot{V}}_{{\rm{water}},{\rm{out}}}\) is the volume yield of liquid water, \({A}\) is the sunlight collection Area (see the approximate part below), \({E}_{{\rm{GHI}}}\) is the GHI in Wsun m−2, and \({\rho }_{{\rm {水} }}\) is the density of water.

This is now a function of three key climate variables: GHI (the first term), ambient temperature (the second term, hidden in the third term), and RH (enter the third term). This is converted into an output table and processed through the AWH-Geo pipeline, and is shown in Figure 3a. Although this can be run for any selected parameter \({T}_{{\rm{hot}}}\), we provide here \({T}_{{\rm{hot}}}\ ) = 100 °C, a temperature that can still be reached in low-cost (non-vacuum) practical equipment without tracking or concentration of sunlight. Higher driving temperature will increase the upper limit of water output. For limit analysis, RH values ​​higher than 90% are clamped to prevent unrealistically high theoretical output when the Kim equation approaches infinity at 100% RH. It is further assumed that the new ambient air is effectively updated.

Figure 3b depicts the active cooler without sensible heat recovery-the maximum output of the condenser-the best coefficient of performance for all given operating inputs and cooling devices at the condenser temperature, as modeled by Peeters32, we learned from their Figure digitized. 11. Peeters chose to set the output to zero when frost is expected to form on the condenser. Since Peeters assumes work input, we converted from solar energy (GHI) to kWhPV as described above.

Figure 3c maps Zhao’s experimental results from TRP to their reported SY using logistic regression curves to fit 0.21, 3.71, and 9.28 l kWh-1 15 at 30, 60, and 90% RH, respectively. The terms of curve fitting are reported in the next section.

The custom yellow to blue map colors are based on www.ColorBrewer.org and the author is CA Brewer of Pennsylvania State University.

Two simple characteristic equations, linear and logical, are used to fit a limited set of SY and RH pairs from laboratory experiments or reported values, and are drawn by AWH-Geo using the calculated output table. A hypothetical curve of similar form, whose terms are iteratively adjusted in AWH-Geo to find the target output (5 l d-1) and user group, and are reported here (for 1 square meter equipment). In the following equation,% RH is considered a fraction (for example, 55% is equivalent to 0.55).

The linear target curve is a simple linear function that intersects the y-axis at zero:

Among them, a is set to 1.60, 1.86, and 2.60 L/kWh to reach the target of 0.5, 1.0, and 2 billion people, respectively, without SMDW, and RH is the input RH (score).

The logical target curve is a logical function:

Where L is set to 1.80, 2.40, and 4.80 L kWh−1 to reach the goal of 0.5, 1.0, and 2 billion people without SMDW, respectively, k is the growth rate set to 10.0, and \({\rm{RH}}\) and \({{\rm{RH}}}}_{0}\) are input RH (score) and 0.60 respectively.

The SY value of TRP reported by Zhao (they call it "SMAG") fits a logical function of the same form with the following parameters: L is set to 9.81 L kWh-1, k is set to 11.25, and RH0 is set to 0.645.

The generated fitting SY configuration file is expanded into an output table. As with all reports that provide SY values ​​instead of complete output tables, this forces us to assume that the heat rate is linear (approximately equal to GHI), which may introduce errors at lower GHI levels. Zhao reports that the SY of TRP materials remains consistent within a temperature range below 40 °C (the material's lower critical dissolution temperature), and performance drops sharply above this temperature. Therefore, we set SY to 0 l kWh−1 in the output table to obtain a temperature ≥ 40 °C.

Bagheri reported on the performance of three existing AWH equipment in a variety of climatic conditions, using the "energy consumption rate" in kWh/L, which can be thought of as the SEC and the simple reciprocal of SY. We did not fit the logistic curve to the reciprocal, but fitted the exponential function to the average SEC of the three devices under the condition of 20 °C or more in the equation:

Where SEC is the specific energy consumption in kWhPV l-1, and RH is the fractional part.

This is applied to RH and as a reciprocal in the output table and run through AWH-Geo. Since Bagheri reported the equivalent of kWhPV, we expanded it to accommodate the GHI input with the above-mentioned photovoltaic conversion efficiency.

For the performance of the ZMW equipment (the company's ~3 m2 SOURCE Hydropanel), we used the value 35 in the panel production contour map in the technical specification table provided on the manufacturer's website. Due to its importance as an early example of SC-AWH products with commercial intent, the decision to include it was made. The value of l per day for each panel is taken at each 10% RH step of 5 kWh m-2, assuming that it represents kWh m-2 d-1, and then divided by 15 kWh (~3 m2 × 5 kWh m- 2) To convert to SY in l kWh−1. According to the generated SY curve, an output table is generated and processed by AWH-Geo.

Given a complete permutation set of RH from 10% to 100% and GHI with a threshold interval of 400 to 700 W m-2, a binary image time series is used to run a coincidence analysis across 70 threshold pairs through AWH-Geo. The average value obtained is multiplied by 24 to indicate that the average hour threshold per day, namely ophd, is met at the same time. The following is the functional representation of this time series calculation:

Where \({{\rm{RH}}})_{t,{\rm{px}}}\) are map pixels\({\rm{px}}\) at time\(t\), \ ({{\rm{RH}}}}_{{\rm{threshold}}}\) is the RH threshold that the device assumes to operate, \({{\rm{GHI}}} _{t,{\rm{ px}}}\) is the GHI in the map pixel\({\rm{px}}\) at time \(t\), and \({{\rm{GHI} }}_{{\rm{threshold }}}\) is the threshold of GHI. If the threshold is exceeded, the device will be assumed to operate.

Then perform population calculations on these images in Google Earth Engine.

Use grouped image reduction (at a 1,000-meter scale) to sum the population integer counts of populations without previously created SMDW distribution images (from WorldPop), and perform regional statistics on the average ophd image as an integer (0-24). This reduction is carried out at 1,000 m. In Google Earth Engine, a single country in a single ophd area was verified, and the results showed that the error was very small (<2%). The population results are collected as a table (feature set), and the population is cumulatively summed in the stacked ophd area. These are exported to R for drawing in Figure 4b.

In order to assess the sensitivity of the results to the choice of climate and population data sets, we performed a coincidence analysis using alternative data sets (Figure 4b) and provided these results in Figure 1 of the extended data.

As an alternative climate data set for ERA-5 (1 hour, 9 kilometers), we used NASA's Global Terrestrial Data Assimilation System (GLDAS) 2.1 during this period, with a spatial resolution of 0.25° × 0.25° (nominal 30 kilometers) And 3 hour time resolution 65 performed simultaneously with the main results, 2010-2019. As an alternative population data set for WorldPop 2017, we used the LandScan 2017 environmental population count of Oak Ridge National Laboratory, with a spatial resolution of 1 km66. Two result comparisons are calculated: (1) GLDAS calculated using WorldPop 2017 is used to directly compare climate data input, and (2) GLDAS calculated using LandScan is used to compare climate and population data set substitutions.

The comparison shows that the sensitivity to the population data set used is negligible, but it has substantial and systematic sensitivity to the climate data set used, and all the comparisons are consistent in the main features and qualitative conclusions. Compared with the finer ERA-5 climate reanalysis, the GLDAS data set, which is coarser in space and time (3x), has consistently lower predictions of water output and impact. We speculate that the 3-hour time step of GLDAS is not sufficient to capture performance-critical humidity and GHI dynamics throughout the day (maybe morning and evening). Similarly, 30 kilometers of pixels are not enough to solve the fine climate model driven by topography and other micro-geographic effects. . This illustrates the importance of using high-resolution climate data sets.

To surpass the annual average and research availability, we introduced a set of indicators that we named the 90th percentile of Moving Average Density (MADP90).

MADP90-t represents the average output rate of the device (l d-1 m-2), which will exceed the average output rate during 90% of t days at a given location. MADP90 is calculated based on the derived P90 value of the probability density function (PDF) of the daily average output during each t-day window in the time series (2010-2019). The result is a scalar that can be mapped spatially. This study examined moving window periods of 1, 7, 30, 60, 90, and 180 days. MADP90-results can be used as additional results and map layers in AWH-Geo.

Expanded Data Figure 3 provides a collection of PDF examples from a location in southwestern Tanzania. Each P90 value corresponds to a version of the MADP90 indicator corresponding to the moving window period. In most geographic locations, the P90 value will naturally increase with t because the PDF will tighten its dispersion with respect to the natural (P50) mean.

The software and data sets generated and/or analyzed during the current study are available in the following repositories. GitHub: https://github.com/AWH-GlobalPotential-X/AWH-Geo; Figshare: https://doi.org/10.6084/m9.figshare.c.5642992.v1; JMP geographic processor package (Python and ArcGIS geoprocessing model); JMP Geofabric data set (shapefile); population without SMDW image data layer (geoTiff); upper limit AWH output data layer (geoTiff); coincidence analysis result data sheet (Sheets); and output used in this study Table (table). This article provides source data.

The software used in the current study is as follows. GitHub: https://github.com/AWH-GlobalPotential-X/AWH-Geo; AWH-Geo application: processor and output viewer with source code; population and result data processing script.

Bain, R., Johnston, R., Mitis, F., Chatterley, C., and Slaymaker, T. Establish a sustainable development goal baseline for household drinking water, sanitation, and personal hygiene services. Water 10, 1711 (2018).

Progress in drinking water, sanitation, and sanitation: 2017 update and SDG baseline (World Health Organization and UNICEF, 2017).

Hanikel, N. etc. Fast circulation and excellent output in the metal organic frame water trap. ACS Center for Science. 5, 1699–1706 (2019).

CAS PubMed PubMed Central Article Google Scholar 

Li, R., Shi, Y., Wu, M., Hong, S. & Wang, P. Increasing atmospheric water production: using nano adsorbents to achieve multiple water collection cycles. Nano Energy 67, 104255 (2020).

Qi, H. etc. The interface solar energy based on the liquid adsorbent drives the atmospheric water generator and simultaneously adsorbs and desorbs. Senior alma mater. 31, 1903378 (2019).

Humphrey, JH, etc. Atmospheric water collection accelerates the potential of households to obtain safe water. The Lancet Planet. Health 4, e91-e92 (2020).

Gido, B., Friedler, E. & Broday, DM evaluate atmospheric moisture collection through direct cooling. atmosphere. Reservoir 182, 156–162 (2016).

Tu, R. & Hwang, Y. Overview of atmospheric water harvesting technology. Energy 201, 117630 (2020).

With the support of billionaires, Gould, T. sells water to the poorest people in the world at a price of $150/m3. Global Water Information Volume. 21 (May 21, 2020).

Kim, H. et al. Adsorption-based atmospheric water collection device for arid climates. Nat. Community. 9, 1191 (2018).

ADS PubMed PubMed Central Article CAS Google Scholar 

LaPotin, A., Kim, H., Rao, SR & Wang, EN Adsorption-based atmospheric water collection: the effect of material and component characteristics on system-level performance. Cumulative Chemical Reservoir 52, 1588–1597 (2019).

CAS PubMed article Google Scholar 

Tu, Y., Wang, R., Zhang, Y. & Wang, J. Progress and expectations of atmospheric water collection. Joule 2, 1452–1475 (2018).

Gorelick, N. etc. Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote sensing environment. 202, 18-27 (2017).

Gleick, PH Basic water demand for human activities: to meet basic needs. Water International 21, 83-92 (1996).

Zhao F. et al. Super absorbent gel for all-weather atmospheric water collection. Senior alma mater. 31, 1806446 (2019).

Kim, H. et al. Use a metal organic frame driven by natural sunlight to collect water from the air. Science 356, 430–434 (2017).

ADS CAS PubMed article Google Scholar 

Matsumoto, K., Sakikawa, N. & Miyata, T. Thermally responsive gel that absorbs moisture and exuded water. Nat. Community. 9, 2315 (2018).

ADS PubMed PubMed Central Article CAS Google Scholar 

A global indicator framework for the Sustainable Development Goals and targets of the 2030 Agenda for Sustainable Development (United Nations, 2018).

Hutton, G. and Varughese, M. The cost of achieving the 2030 Sustainable Development Goals on Drinking Water, Sanitation and Sanitation (Water and Sanitation Plan, 2016).

Brown, J. and Clasen, T. To obtain health benefits from water quality interventions, a high degree of persistence is necessary. PLoS ONE 7, e36735 (2012).

ADS CAS PubMed PubMed Central Article Google Scholar 

Cohen, A. & Ray, I. Increasingly rely on the global risks of bottled water. Nat. maintain. 1, 327–329 (2018).

Wahlgren, RV Atmospheric water vapor processor design for drinking water production: an overview. Water Resources 35, 1-22 (2001).

CAS PubMed article Google Scholar 

Zhou, X., Lu, H., Zhao, F. & Yu, G. Atmospheric Catchment: A Review of Material and Structural Design. ACS alma mater. Wright. 2, 671–684 (2020).

Domen, JK, Stringfellow, WT, Camarillo, MK & Gulati, S. Fog water as alternative and sustainable water resources. Clean technology. environment. Policy 16, 235–249 (2014).

Kim, J., Park, K., Yang, DR and Hong, S. Comprehensive review of energy consumption in seawater reverse osmosis desalination plants. Application Energy 254, 113652 (2019).

Pullan, RL, Freeman, MC, Gething, PW & Brooker, SJ Improving geographic inequality in drinking water supply and sanitation in sub-Saharan Africa: mapping and spatial analysis of cross-sectional survey data. Public Science Library Medicine. 11, e1001626 (2014).

PubMed PubMed Central article Google Scholar 

Deshpande, A. etc. Map the geographic inequalities of access to drinking water and sanitation in low- and middle-income countries from 2000-17. Lancet ball. Health 8, e1162–e1185 (2020).

Mujwahuzi, MR etc. Drawers of Water II: 30 years of changes in East African household water use and environmental health. (International Institute for Environment and Development, 2002).

Roche, R., Bain, R. & Cumming, O. A long way to go-an estimate of comprehensive coverage of water, sanitation, and hygiene in 25 sub-Saharan African countries. PLoS ONE 11, e0171783 (2017).

Beysens, D. Estimated global dew production based on some meteorological data. atmosphere. Reservoir 167, 146–155 (2016).

Mulchandani, A. & Westerhoff, P. Geospatial climate factors affect the water production of atmospheric water capture devices driven by solar desiccants. environment. science. technology. 54, 8310–8322 (2020).

ADS CAS PubMed article Google Scholar 

Peeters, R., Vanderschaeghe, H., Rongé, J. & Martens, JA Energy performance and climate dependence of freshwater production technology from atmospheric water vapor. environment. science. Water technology. 6, 2016-2034 (2020).

Kim, H., Rao, SR, LaPotin, A., Lee, S. & Wang, EN Thermodynamic analysis and optimization of atmospheric water collection based on adsorption. internationality. J. Heat Transfer 161, 120253 (2020).

Bagheri, F. Performance investigation of atmospheric water collection system. Water Resources Industry 20, 23-28 (2018).

What is the source? https://www.source.co/wp-content/uploads/2020/11/SOURCE-Tech-Spec-Sheet.pdf (accessed in June 2021).

Wang, JY, Wang, RZ, Wang, LW & Liu, JY An efficient semi-open system for producing fresh water from the atmosphere. Energy 138, 542–551 (2017).

Hanikel, N., Prévot, MS & Yaghi, OM MOF water collector. Nature Nanotechnology 15, 348–355 (2020).

ADS CAS PubMed article Google Scholar 

LaPotin, A. etc. Two-stage atmospheric water collection device for scalable solar-driven water production. Joule 5, 166–182 (2021).

Terzis, A. etc. High frequency water vapor adsorption cycle using metal organic framework fluidization. Cells represent physics. science. 1. 100057 (2020).

Logan, MW, Langevin, S. & Xia, Z. Reversible atmospheric water collection using metal organic framework. science. Representative 10, 1492 (2020).

ADS CAS PubMed PubMed Central Article Google Scholar 

Yilmaz, G. et al. Autonomous atmospheric seepage MOF matrix. Sci Adv 6, eabc8605 (2020).

ADS CAS PubMed PubMed Central Article Google Scholar 

Foster, T. & Willetts, J. Multiple water use in rural Vanuatu: Does the family choose the safest way to drink? internationality. J. Environment. Health Resources 6, 579–589 (2018).

Elliott, M. et al. Solve how to build water resilience and support sustainable development for multiple household water sources and uses. npj Clean Water 2, 6 (2019).

Oki, T. Global hydrological cycle and world water resources. Science 313, 1068–1072 (2006).

ADS CAS PubMed article Google Scholar 

Elhenawy, SEM, Khraisheh, M., AlMomani, F. & Walker, G. The metal organic framework serves as a platform for CO2 capture and chemical processes: CO2 adsorption, membrane separation, catalytic conversion and electrochemical reduction. Catalyst 10, 1293 (2020).

DeSantis, D. etc. Technical and economic analysis of metal-organic frameworks for hydrogen storage and natural gas. Energy Fuel 31, 2024-2032 (2017).

Mohapatra, S. and Simon, L. Bargaining within the family regarding home technology adoption. Pastor economy. Family 15, 1263–1290 (2017).

Barstow, CK, etc. Design and pilot projects to provide water filters and improved stoves in Rwanda. PLoS ONE 9, e92403 (2014).

ADS PubMed PubMed Central Article Google Scholar 

Daniel, D., Marks, SJ, Pande, S. & Rietveld, L. Social and environmental drivers for the sustainable adoption of household water treatment in developing countries. npj Clean Water 9, 12 (2018).

Smits, J. & Permanyer, I. Local Human Development Database. science. Data 9, 190038 (2019).

GADM v3.6. Global administrative regions in 2020 (University of California, Berkeley, 2020); https://gadm.org

Spatial Data Repository, Demographic and Health Survey Program (USAID, 2020) https://spatialdata.dhsprogram.com

GDL v3.7.0. Global Data Lab Regional Database 2020 (Radboud University Management Institute, 2020); Global Data Lab

Incorporating water quality testing into household surveys: a special report on drinking water (UNICEF and World Health Organization, 2020).

Bain, R., Johnston, R., Khan, S., Hancioglu, A. & Slaymaker, T. Monitoring drinking water quality in a nationally representative household survey in low- and middle-income countries: a cross-section of 27 multiple indicators A face-to-face analysis of the 2014-2020 cluster survey. environment. Health perspective. 129, 97010 (2021).

Lloyd, CT, etc. A global spatiotemporal coordination dataset used to generate a high-resolution grid population distribution dataset. Big Earth Data 3, 108–139 (2019).

PubMed PubMed Central article Google Scholar 

Bain, R., Johnston, R. and Slaymaker, T. Drinking water quality and sustainable development goals. npj Clean Water 3, 37 (2020).

Jacobson, MZ and Jadhav, V. World estimated PV best tilt angle and the ratio of sunlight incident on tilted and tracked PV panels relative to horizontal panels. Solar 169, 55–66 (2018).

Champion photovoltaic module efficiency chart. National Renewable Energy Laboratory https://www.nrel.gov/pv/assets/pdfs/champion-module-efficiencies.20200708.pdf (2020).

Muñoz Sabater, J. ERA5-Hourly land data from 1981 to the present. Copernicus Climate Change Service (C3S) Climate Data Storage (CDS) https://doi.org/10.24381/cds.e2161bac (2019).

The modified Magnus form of Alduchov, OA and Eskridge, RE approximates the saturated vapor pressure. J. Application Meteorology. climate. 35, 601–609 (1996).

Iowa Environmental Mediation (Iowa State University, 2020); https://mesonet.agron.iastate.edu/

Blanco, MJ, Martı ́n, JG & Alarcón-Padilla, DC Theoretical efficiency of angle-selective non-concentrating solar thermal systems. Solar Power 76, 683–691 (2004).

Brewer, California ColorBrewer2. http://www.ColorBrewer.org (2020).

Beaudoing, H., Rodell, M. GLDAS Noah Land Surface Model L4 0.25 × 0.25 degrees per month V2.1 (Goddard Earth Science Data and Information Service Center, 2020); https://doi.org/10.5067 /SXAVCZFAQLNO

Rose, AN, McKee, JJ, Urban, ML and Bright, EA LandScan 2017 (Oak Ridge National Laboratory, 2018).

We thank many colleagues for their contributions, including A. Aron-Gilat, D. Youmans, GL Whiting, M. Eisaman, S. Lin, J. Sargent, S. McAlister, S. Chariyasatit, B. Dixon, E. St Jean Duggan , F. Carlsvi, K. Stratton, M. McCoy, R. Hessmer, J. Hanna, H. Riley, P. Watson, M. Day, B. Quintanilla-Whye, A. Ramadan, A. Little and D. Moufarege We thank the WHO/UNICEF JMP team for their guidance on drinking water service estimates, especially T. Slaymaker, R. Johnston and F. Mitis; the Google Earth Engine team, especially S. Ilyushchenko, S. Agarwal, T. Erickson, N. Gorelick, M. Hancher, M. Dixon, M. DeWitt, J. Conkling, N. Clinton, K. Reid, E. Engle, W. Rucklidge and the entire Earth Engine development community for advice; C. For code review Caywood; B. Schillings and J. Gagne are sponsored internally by X. Funding was provided by Google LLC.

X, The Moonshot Factory, Mountain View, California, USA

Jackson Rhodes, Ashley Thomas, Neil Tritt, Matthew Foggin, Cyrus H. Beluz, Tillek Mamutov, Gillian Fonheiser, Nicole Kobylansky, Shane Washburn, Claudia Trussdale, Claire Lee, and Philip H. Schmeltz

WHO/UNICEF Joint Monitoring Program, Department of Data, Analysis, Planning and Monitoring, UNICEF, New York, New York, USA

Google Inc., Mountain View, California, USA

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

PHS and JL conceived this research. JL, PD, TM and NT were analyzed and graphed. AT, NT, JL, PHS, RB and CHB made their points. JL, PHS, AT and RB wrote this paper. This research is a subset of X’s greater efforts, led by PHS, MF, NT, and AT, and aims to develop household AWH as a commercial product, providing information for current research: MF, NT, and SW led prototype development and experiments , CHB conducts physical modeling, MF, SW, CT, CL, etc. build equipment and conduct experiments, AT, JF and NK conduct market and user research.

Correspondence with Jackson Lord or Philipp H. Schmaelzle.

We disclose the following potential competing interests. This work was funded by X, The Moonshot Factory (previously known as Google[x]). X is part of Alphabet. Both are for-profit entities. X has applied for patent protection for the water-in-air device, and many of the authors are listed as inventors. After meeting certain indicators, air-to-water equipment may represent an important business opportunity. This work can be further carried out in various ways, including as a possible spin-off company, in which one or more authors may become founders, managers, shareholders, employees or otherwise involve economic interests.

Peer review information Nature thanks Marisa Escobar and other anonymous reviewers for their contributions to the peer review of this work. Peer review reports are available.

The publisher states that Springer Nature remains neutral on the jurisdiction claims of published maps and agency affiliates.

Compared with the main results of the compliance analysis of the ERA5-Land and WorldPop 2017 datasets (Figure 4b, people without SMDW are served by opH/d that meets the climate threshold) and (a) the results of GLDAS 2.1 climate and WorldPop 2017 population, And (b) GLDAS 2.1 climate and LandScan 2017 population data set. Shows the number of operating hours per day (opH/d) for global horizontal irradiance (GHI) and relative humidity (rH) thresholds.

(a) Charts and (b) Tables to verify the use of known reference data sets to cross-estimate the percentage of safety management (SM) from the national (N) breakdown at least from the sub-national (SN) level of the basic (ALB) drinking water ladder SN level from WHO/UNICEF JMP data. The reference value of the nationally representative multi-index cluster survey combined with water quality testing (reference SM) is compared with the estimated value of JMP Geoprocessor combined with the estimated value of JMP subnational ALB and the estimated value of national safety management drinking water service (est. SM). Ordinary least squares regression (OLS) results in the reported standard error (stdErr). Sample size n = 15. Table (b) shows the main result (ERA5-Land) population count after regression adjustment. Global horizontal irradiance (GHI) and relative humidity (rH) thresholds indicate a population without safely managed drinking water (SMDW).

Histogram of moving average production (L/d/m2) across a window period (in days) at a location in Manda, Tanzania. The P90 availability value increases as the average window period increases. The P90 value is an estimate and is for illustrative purposes only.

(a) MADP90-90day and (b) MADP90-7day (a measure of availability over time) during the 10-year analysis period 2010-2019 (inclusive), the global AWH thermodynamic upper limit (Kim 2020).

Biweekly average output (L/d/m2) and climate input global horizontal irradiance (GHI, plotted from 0–1000 W/m2), relative humidity (rH, plotted from 0–100%) and temperature (from plot ( a) (a) AWH thermodynamic upper limit (Kim 2020) of 0-100 °C) at every two-week interval between 2010-2019 (inclusive) and (b) averaged once every two weeks during this period, (c) A linear target curve for 1 billion users in each bi-weekly interval. Examples of stable, low-variability output profile characteristics in equatorial tropical regions.

Biweekly average output (L/d/m2) and climate input global horizontal irradiance (GHI, plotted from 0–1000 W/m2), relative humidity (rH, plotted from 0–100%) and temperature (from plot ( a) (a) AWH thermodynamic upper limit (Kim 2020) of 0-100 °C) at every two-week interval between 2010-2019 (inclusive) and (b) averaged once every two weeks during this period, (c) A linear target curve for 1 billion users in each bi-weekly interval. An example of a seasonal wet and dry savanna climate, where the AWH output driven by rH fluctuates moderately for half a year.

The biweekly average output (L/d/m2) and climate input global horizontal irradiance (GHI, drawn from 0–1000 W/m2), relative humidity (rH, drawn from 0–100%) and temperature (drawn from (a) (a) AWH thermodynamic upper limit (Kim 2020) of 0-100 °C) at every two-week interval during the ten-year period from 2010 to 2019 (inclusive) and (b) averaged once every two weeks per year during this period, (c ) A linear target curve of 1 billion users in each bi-weekly interval. As an example of a seasonal wet and dry savanna climate, the AWH output driven by rH fluctuates significantly half a year.

Biweekly average output (L/d/m2) and climate input global horizontal irradiance (GHI, plotted from 0–1000 W/m2), relative humidity (rH, plotted from 0–100%) and temperature (from plot ( a) (a) AWH thermodynamic upper limit (Kim 2020) of 0-100 °C) at every two-week interval between 2010-2019 (inclusive) and (b) averaged once every two weeks during this period, (c) A linear target curve for 1 billion users in each bi-weekly interval. The temperature-driven AWH outputs an example of a mid-latitude climate that fluctuates significantly in half a year.

(a) The overall average output (L/d/m2) of the target logistic curve for the 1 billion users of 5 L/d/m2 during the 10-year period from 2010 to 2019 (inclusive). (b) The average output ratio (%) of the same specific output (SY, in L/kWh) distribution during the ten years from 2000 to 2009 (inclusive) is abnormal. Red indicates that AWH output has increased over time between two decades. Blue means the AWH output is reduced.

Open Access This article has been licensed under the Creative Commons Attribution 4.0 International License Agreement, which permits use, sharing, adaptation, distribution and reproduction in any media or format, as long as you appropriately indicate the original author and source, and provide a link to the Creative Commons license , And indicate whether any changes have been made. The images or other third-party materials in this article are included in the article’s Creative Commons license, unless otherwise stated in the material’s credit line. If the article’s Creative Commons license does not include the material, and your intended use is not permitted by laws and regulations or exceeds the permitted use, you need to obtain permission directly from the copyright owner. To view a copy of this license, please visit http://creativecommons.org/licenses/by/4.0/.

Lord, J., Thomas, A., Treat, N. etc. The global potential of using solar energy to collect drinking water from the air. Nature 598, 611–617 (2021). https://doi.org/10.1038/s41586-021-03900-w

DOI: https://doi.org/10.1038/s41586-021-03900-w

Anyone you share the following link with can read this content:

Sorry, there is currently no shareable link in this article.

Provided by Springer Nature SharedIt content sharing program

By submitting a comment, you agree to abide by our terms and community guidelines. If you find content that is abusive or does not comply with our terms or guidelines, please mark it as inappropriate.

Nature ISSN 1476-4687 (online) ISSN 0028-0836 (print)