How Much Atmospheric Water Can We Harvest Before Local Humidity Changes?
- Amin Mojiri
- 11 minutes ago
- 9 min read
Original Authors: Carl Abadam, Anjali Mulchandani
Original paper is accessible at: https://doi.org/10.1021/acsestengg.6c00532
A new question for atmospheric water harvesting
Atmospheric water harvesting (AWH) is often presented as a promising route for producing water from an abundant atmospheric reservoir. The atmosphere contains approximately 12,900 km³ of water, about six times the volume of freshwater stored in rivers and freshwater lakes. This makes AWH attractive as a decentralized water source, especially for communities with limited access to conventional surface water, groundwater, or seawater supplies.
Most AWH research has focused on materials, devices, energy demand, technoeconomic feasibility, and water quality. This paper asks a different and less-examined question:
پ
If AWH removes water vapor from air, to what extent can it influence local humidity?
This question matters because the atmosphere is globally large, but water extraction is local. A small household AWH device may have negligible influence, but future large-scale or dense deployments could raise questions about microclimate effects, atmospheric water rights, and responsible environmental governance.
From water production to humidity impact
The authors frame AWH through a simple but powerful analogy: water withdrawal from a river.
When water is withdrawn from a river, downstream flow decreases. Similarly, when AWH extracts water vapor from air, the amount of water vapor remaining in the processed control volume decreases.
The paper expresses this idea through a mass balance:
post-AWH atmospheric moisture = starting atmospheric moisture − water extracted by AWH
In other words, AWH does not create new water; it diverts water vapor from the atmospheric reservoir into liquid water.
The paper uses this mass-balance framework to evaluate:
RHstart: starting relative humidity
RHend: post-AWH relative humidity
QAWH: water vapor extracted by AWH
BV: buffer volume of surrounding air
k: spatial scaling factor
θ: moisture retention ratio
The central idea is that AWH impact depends not only on how much water is harvested, but also on the volume of air over which that extraction is diluted.
The buffer volume concept
A major contribution of the paper is the concept of a buffer volume (BV).
The buffer volume represents the surrounding air volume available to dilute the humidity reduction caused by AWH. If a device removes water vapor from a small, confined room, the local humidity drop can be significant. If the same extraction is distributed across a much larger outdoor air volume, the modeled humidity change becomes much smaller.
The processed control volume is defined as:
V = Q × t
where:
Q is the volumetric air flow rate of the AWH device
t is the harvesting duration
For example, a device with Q = 300 m³/h operating for one hour processes 300 m³ of air. The total control volume is then expanded by the spatial scaling factor k, giving larger volumes such as 600 m³ at k = 2 and 900 m³ at k = 3.
This allows the authors to evaluate how the same AWH extraction is diluted across increasingly large air volumes.
A conservative upper-bound model
The model assumes a theoretical 100% harvest efficiency as an upper-bound case. In this scenario, all water vapor is removed from the processed control volume, while the remaining surrounding air volumes are assumed to stay at the starting vapor density.
This is intentionally conservative. It does not simulate a fully dynamic atmosphere; rather, it provides a closed-volume, conservation-of-mass estimate for how much humidity could change under simplified conditions.
Because relative humidity is linearly proportional to vapor density at fixed temperature, the authors use the mass balance to estimate RHend from RHstart and the spatial scaling factor k. Their sensitivity analysis shows that changing temperature changes vapor density values, but does not change the resulting relative humidity decrease calculated by the model.
Three climate scenarios
The authors evaluated three representative starting humidity conditions:
25% RHstart: arid climate
50% RHstart: temperate climate
75% RHstart: tropical climate
They modeled post-AWH humidity across spatial scaling factors from k = 1 to 100. When k = 1, the processed control volume is fully dehumidified, so RHend = 0%. As k increases, the same dehumidified air volume is diluted into a larger total control volume, and RHend approaches RHstart.
The relationship is nonlinear. The largest humidity changes occur at small values of k. For example, at RHstart = 75%, the absolute humidity decrease is:
75 percentage points at k = 1
37.5 percentage points at k = 2
8.3 percentage points at k = 9
7.5 percentage points at k = 10
This shows that the influence of AWH is strongest when the surrounding buffer volume is small. As the spatial scale increases, the modeled relative humidity change becomes progressively smaller.
AWH impact is scale-dependent
One of the paper’s most important messages is that AWH influence is strongly scale-dependent.
At small spatial scales, such as rooms, warehouses, or poorly ventilated spaces, AWH could meaningfully reduce local humidity because the available buffer volume is limited.
At larger outdoor scales, the modeled influence becomes much smaller as the buffer volume increases. In real outdoor environments, the effect is expected to be further diluted by large accessible air volumes and atmospheric mixing; however, the present model treats this through buffer volume rather than dynamic airflow simulation. The authors explicitly note that their results are not intended to minimize small-scale moisture removal, but to place it in context at larger spatial extents.
What happens at 1 MGD?
The paper provides an important numerical example for large-scale AWH.
For an AWH system harvesting:
1 million gallons per day (MGD)
within a buffer volume of:
10 km³
corresponding to a 1 × 1 km² area extending upward to a 10 km tropospheric height
the model predicts that relative humidity decreases from:
RHstart = 50%
to RHend = 47.8%
If the same 1 MGD harvest is spread across a buffer volume 25× larger:
250 km³
corresponding to 5 × 5 × 10 km³
the modeled humidity decreases only to:
RHend = 49.9%
The authors also report that extraction would need to increase by about four orders of magnitude, to >10,000 MGD, to completely remove humidity in the larger buffer volume.
This result is central to the paper:
AWH can influence local humidity, but the magnitude depends strongly on how much air volume is available to dilute the extraction.
Buffer volume requirements increase with water production
The paper also estimates how much surrounding air is needed to maintain a desired moisture retention ratio after AWH.
For RHstart = 50% at 20°C, with starting water vapor density of 0.0087 kg/m³, and a harvest rate of:
100 MGD
the required buffer volume depends strongly on the target moisture retention ratio:
θ = 0.99 requires 181 km³ of air
θ = 0.95 requires 36.3 km³
θ = 0.90 requires 18.1 km³
θ = 0.80 requires 9.06 km³
θ = 0.50 requires 3.63 km³
This means that maintaining near-ambient humidity while harvesting large amounts of water requires either:
reducing the water extraction rate, or
increasing the effective atmospheric buffer volume over which extraction is distributed.
The authors note that outdoor systems may benefit from large accessible air volumes and atmospheric mixing, while indoor or confined AWH systems are more likely to create stronger local humidity depletion unless additional airflow or humidity control is used.
Device spacing and deployment density matter
Another practical contribution of the paper is its discussion of device spacing.
The authors approximate the buffer volume as a rectangular column extending to the troposphere. They use a characteristic height of:
10 km
or 10,000 m
Based on this framework, the spacing between AWH devices can be estimated so that their buffer volumes do not overlap. For example, a 10 km³ buffer volume corresponds to a horizontal footprint of approximately 1 km × 1 km, giving a characteristic spacing of about 1 km between devices under the simplified model.This matters because if devices are placed too close together, partially dehumidified air from one device could be processed again by another device.
The paper describes a worst-case “in-series” scenario in which moisture retention compounds across sequential buffer volumes. For example:
θ = 0.95 becomes approximately 0.90 after two sequential buffer volumes
θ = 0.95 becomes approximately 0.86 after three sequential buffer volumes
This highlights the importance of device spacing, deployment density, and network configuration when designing future AWH systems.
Current AWH systems are far below climate-impact thresholds
A key reassuring conclusion is that present-day AWH devices are far too small to measurably affect surrounding climate at broad environmental scales.
The authors modeled harvesting rates from 10⁻³ to 10⁶ MGD across three climate scenarios and several buffer volumes. For a large metropolitan-scale buffer volume of 4000 km³, completely reducing relative humidity to RHend = 0% under arid conditions would require an unrealistically large extraction rate on the order of 10⁵ MGD. The paper uses Albuquerque, New Mexico, with an approximate spatial footprint of 400 km², as a reference for this scale.
By comparison, currently available commercial AWH systems advertise production up to:
3000 L/day
approximately 0.00079 MGD
These systems are several orders of magnitude below the threshold required to induce measurable local humidity change. Therefore, under current deployment scales, the influence of AWH on the surrounding climate is effectively negligible.
The paper also notes that large municipal water treatment plants typically supply water on the order of >150 MGD. Even if AWH were hypothetically used to replace such water production, the resulting humidity change would remain negligible when distributed across sufficiently large effective buffer volumes.
Why this matters for environmental governance
Although current AWH systems are unlikely to alter local climate, the authors argue that early quantitative assessment is important before the technology scales.
They compare the issue to other water-resource histories where assumptions of abundance led to unintended consequences. For example, Albuquerque once relied heavily on groundwater that was thought to be nearly limitless. Later studies showed the aquifer was smaller than believed; groundwater was being withdrawn at roughly twice its natural replenishment rate, and water tables had dropped by as much as 120 feet in parts of the city. Recovery required new infrastructure to divert San Juan–Chama surface water for part of the year, a project costing $400 million.
The authors also point to the Colorado River Compact, which originally assumed 16.5 million acre-feet (MAF) of natural flow, while later calculations suggest the average is closer to 12.4 MAF. This historical overestimation contributed to long-term allocation conflict.
These examples are not direct analogies to AWH, but they support the paper’s broader message:
Resource governance should begin before a resource becomes contested.
Who owns water in the air?
The paper raises an important question:
Who owns atmospheric water vapor?
Surface water and groundwater are regulated, but atmospheric moisture is not yet clearly defined as a harvestable water resource. The authors discuss possible legal analogies, including riparian doctrine, prior appropriation, rainwater harvesting, and cloud seeding.
Rainwater harvesting is already regulated differently across states. For example, the paper notes that Colorado limits rainwater storage to 110 gallons, or 416 L.
The authors do not claim to solve atmospheric water rights. Instead, they ask questions that future policy may need to address:
Can landowners claim atmospheric water vapor above their property?
Could the first large-scale AWH user claim a water right?
Should there be caps on atmospheric water extraction?
Would interstate or intergovernmental agreements be needed for large-scale AWH deployment?
The purpose is to begin a scientific and policy conversation before atmospheric water becomes vulnerable to competition or depletion.
Limitations and future research
The authors clearly describe the model as a first-order framework. It does not fully simulate:
wind speed
turbulent mixing
atmospheric stability
advection
dynamic replenishment
terrain effects
real device efficiency variations
Instead, it provides an initial conservative mass-balance framework for estimating order-of-magnitude impacts. Future models could incorporate geospatial mapping, field validation, and humidity sensors placed at different distances from AWH devices to measure real microclimate effects.
This makes the paper valuable not because it predicts every real-world scenario, but because it creates a quantitative starting point for responsible AWH deployment.
Key insights
AWH removes water vapor locally, even though the atmospheric reservoir is globally large.
The atmosphere contains about 12,900 km³ of water, approximately six times the volume stored in freshwater rivers and lakes.
The environmental influence of AWH depends on both:
amount of water harvested
atmospheric buffer volume available for dilution
At 1 MGD in a 10 km³ buffer volume, the model predicts RH decreasing from 50% to 47.8%.
The same 1 MGD spread across 250 km³ reduces RH only to 49.9%.
At 100 MGD, maintaining 99% moisture retention requires about 181 km³ of air.
Current commercial systems producing up to 3000 L/day (~0.00079 MGD) are far below thresholds expected to measurably affect local humidity.
Dense AWH deployments could create overlapping dehumidification zones if device spacing and air mixing are not considered.
Indoor or confined AWH systems may have stronger microclimate effects than outdoor systems.
The paper opens an important conversation about atmospheric water rights, environmental monitoring, and responsible deployment.
Takeaway
This study does not argue that atmospheric water harvesting is environmentally harmful at today’s deployment scales.
Instead, it provides a framework for asking the right question before the technology scales further:
How much water can we remove from air before local humidity meaningfully changes?
The answer depends on scale. Small current devices are unlikely to matter climatically. Large, dense, or confined AWH deployments deserve more careful modeling, spacing, monitoring, and governance.
The most important contribution of this work is shifting AWH evaluation beyond yield, energy, and cost toward a broader question of environmental footprint and atmospheric water stewardship.
Comments