Heat, Drought and Our Water Supply Future

Is Hot Drought a Risk in the US Mid‐Atlantic? A Potomac Basin Case Study - Schultz - 2025 - JAWRA Journal of the American Water Resources Association - Wiley Online Library

C. L. Schultz, A. Seck, S. N. Ahmed; Is Hot Drought a Risk in the US Mid-Atlantic? A Potomac Basin Case Study ; 18 June 2025 https://doi.org/10.1111/1752-1688.70031

The Potomac River is a major source of water for our region and the only source of water for Washington, D.C. and Arlington, VA. Over 6 million people in Virginia, Maryland, Pennsylvania, West Virginia, and Washington DC and the diverse ecosystems of the interstate area depend on the water resources of the Potomac River basin. Responsible management of this resource is necessary to ensure all our futures.

The Interstate Commission on the Potomac River Basin (ICPRB) was created to protect and enhance the waters and related resources of the Potomac River basin through science, regional cooperation, and education. The ICPRB has no regulatory authority, but has teamed up with state, federal, and local agencies, and private citizens to work towards a plan that will serve as a roadmap for the sustainable use of this interstate resource now and into the future.

A recently published article by Dr. Cherie Schultz and colleagues in ICPRB’s Section for Cooperative Water Supply Operations on the Potomac cited above, explores the future risks and impacts of hot drought through the lens of the Potomac River basin. In the study which I have excerpted from freely below, the researchers explain why estimating future water requirements in the region is a challenge, but in the face of increased variability in precipitation and temperatures it is essential for the sustainability of our region.

As we prepare for the impacts of climate change (that we can no longer stop), we must plan for adequate supplies of fresh water. Water is essential for human life and well-being—necessary for households for drinking, cooking, and sanitation, for industries, energy producers, and agricultural producers—and for aquatic life that depends on freshwater ecosystems. Rivers and streams are a primary source of water for human use especially in urban areas like the DMV.

We size reservoirs and other infrastructure to meet the demands of anticipated severe droughts and storms. Extreme events are key drivers in the development of water management strategies. It can be expensive or catastrophic to guess wrong.  Thus, it is essential to estimate changes and variability in stream and river flow to adequately assess future infrastructure needs.

Water supply planning, which relies on climate projections at the regional scale, involves considerable challenges. First and foremost is the lack of confidence in general climate model (GCM) projections of regional precipitation, the primary driver of streamflow projections. But for a given region, different GCMs can give very different precipitation projections, resulting in wide ranges of estimates for future streamflow, not uncommonly including disagreements on whether to expect an increase or decrease in stream flow.  (Green Risks: Climate Models Biases and Limits). New infrastructure to mitigate the impacts of climate change on water supply systems can be very costly and hydrologic modeling studies based on a limited number of climate projections may not be enough to compel decision makers to move forward and commit tax dollars.

Other researchers (Zhang et al. (2023)) have argued that measured/observation-based statistically derived climate response functions provides advantages over earth systems models in the simulation of future stream flows. A major uncertainty in these simulations is the response of streamflow to changes in temperature. Rising temperatures will affect streamflow by increasing evaporative losses from soils and water surfaces and by changing the timing and magnitude of snowmelt. Hydrological droughts may become more severe due to increasing temperatures coupled with natural variability in precipitation and result in extreme events which have been characterized as “hot drought.” (Udall and Overpeck 2017; Woodson et al. 2021).    

In this study, the researchers at the ICPRB created an approach for estimating future trends in annual streamflow for the Potomac River. To project flows in extreme drought years, they projected changes in the distribution function of annual streamflow under a future climate and using machine learning to perform statical modeling to estimate the missing data  to create a time series, of sufficient length to compute extreme percentile values (100 year  droughts and 100 year floods), are created by pooling shorter time series from multiple GCMs.

Flow in the freshwater portion of the Potomac River is measured at the US Geological Survey (USGS) stream gage at Little Falls Dam near Washington, DC  located below the intakes of the metropolitan area water suppliers and a above the head of tide in the Potomac estuary. With few major impoundments in the 11,560 square mile drainage area above Little Falls Dam, river flow is largely unregulated and highly variable. Some degree of storage is provided by the underlying fractured bedrock groundwater aquifers, but baseflow recession rates are typically on the order of months, so this storage can be rapidly depleted during periods of low precipitation (Schultz et al. 2014). Precipitation above Little Falls averages 40.3 inches annually, with evapotranspiration averaging 65%. Precipitation is fairly uniform throughout the year, but river flow exhibits a pronounced seasonal variation due to higher evapotranspiration rates during the March–September growing season, which reduce both groundwater recharge and runoff (Trainer and Watkins 1975). Flow tends to be highest in the month of March, with a long-term mean of 1,421,768 cubic feet per minute, and lowest in September, with a long-term mean of 233,076 cubic feet per minute. The snowpack that may accumulate at higher elevations during the winter months slightly increases median river flows in March and April but does not persist long enough to have a significant impact on summertime flows (Cummins et al. 2010).

Study results for the Potomac are consistent with past findings that precipitation will increase in the Mid-Atlantic region as temperatures rise (Shenk et al. 2021), but also indicate that decreases in Potomac River flows and resulting reductions in water supply availability may be experienced in some years. Median temperature is projected to increase by 2.3° to 3.2°C, from the baseline period, 1950–1979, to the planning period, 2040–2069, and by 2.0° to 4.8°C in 2080–2099, and median precipitation is projected to increase by 9%–12% and 11%–16% over the same time periods.

Applying statistical modeling, the data indicate that, “future Potomac River flows will be impacted by ‘hot drought’, that is, increasing drought severity caused by rising temperatures coupled with natural variability in precipitation.” Even though precipitation amounts are expected to increase by up to 16% by 2099, annual river flows may decrease by as much as 49% by the same year due to extreme heat. The increased rain does not mitigate the conditions in the extremely dry and hot months.  The scientists modeling projected river flow changes of −3% to −26% by 2040–2069 and −2% to −49% by 2070–2099 which is a wide range. These results can inform planning, but need a higher degree of certainty before they can be depended  on for investment decisions for the Washington, DC, metropolitan area's cooperative regional water supply system. In our region, assessments of water supply system reliability are largely driven by the impacts of conditions comparable to a 100-year drought. 

Water as a path for Adaptation to Climate Change

 COP29 Declaration on Water for Climate Action

Last week June 16^th-20^th 2025 there were more Climate Conferences. I’m not sure mankind is making progress, it is hard to see in real time. The 62nd Sessions of the Subsidiary Bodies of the United Nations Framework Convention on Climate Change (SB62) took place in Bonn, Germany. They planned to discuss elaboration and revision of National Determined Contributions (NDCs) and National Adaptation Plans (NAPs). There has been very little progress on that, but Friday’s secession was devoted to water.

Friday’s session consisted of two moderated panel discussions and was titled: “Mainstreaming water into NDCs and NAPs for effective adaptation, mitigation and resilience.”

Mainstreaming water into Nationally Determined Contributions (NDCs), National Adaptation Plans (NAPs), and global climate processes to enhance adaptation and mitigation measures. They planned to do this by highlighting related tools and experiences of organizations on how they integrate water, ecosystems, water supply and sanitation. Addressing the transboundary nature and issues of water.

The event was also intended to contribute to the implementation of the Baku Dialogue on Water for Climate Action and the UN System-wide Strategy on Water and Sanitation. This event was organized within the framework of the UN-Water Expert Group on Water and Climate Change, co coordinated by UNESCO, WMO and UNECE, and in cooperation with Azerbaijan and UNEP as partners under the Baku Dialogue on Water for Climate Action, as well as with the Arizona State University (home for Professor Jay Famiglietti), the Association of Sustainable Ecological Engineering Development, and India Water Foundation. This meeting only caught my attention because of the participation of Dr. Famiglietti as the moderator of one of the panels.

Dr Famiglietti is currently a Global Futures Professor in the School of Sustainability at Arizona State University and serves as the Director of Science for the Arizona Water Innovation Initiative. He is Professor Emeritus from the University of Saskatchewan, where he was Executive Director of the Global Institute for Water Security. Before that he served as the Senior Water Scientist at the NASA Jet Propulsion Laboratory at the California Institute of Technology. From 2013 through 2018, he was appointed by Governor Jerry Brown to the California State Water Boards in the Santa Ana and Los Angeles regions. Unfortunately for me, the event was not available to attend remotely, and I certainly was not going to fan girl in Bonn.

The Event was promoted to:  

Serve as a platform to connect water and climate experts and to support countries in integrating and subsequently implementing water-related measures in NDCs and NAPs as one of the priorities of the UN System-wide Strategy on Water and Sanitation;

Contribute to the Baku Dialogue on Water for Climate Action by providing an opportunity for countries that have already endorsed the COP29 Declaration on Water for Climate Action to exchange experiences and to encourage other countries to endorse the declaration and to join the initiative.

I asked CoPilot to   Generate a Summary of the COP29 Declaration on Water for Climate Action and then I edited it down because it was too verbose. The COP29 declaration highlights the critical importance of water in addressing climate change, emphasizing its central role in the impacts and solutions for climate challenges.

Key Concerns:

• Water is significantly affected by climate change, with impacts manifesting through floods, droughts, water scarcity, and other alterations in the water cycle.
• There is an urgent need for protecting, conserving, and restoring water resources and ecosystems, such as rivers, lakes, and groundwater, for effective climate mitigation and adaptation.
• Globally, 2.2 billion people lack access to safe drinking water, while 3.5 billion lack safely managed sanitation. Half the population faces severe water scarcity.
• Water-related disasters, including floods, are responsible for over 90% of disaster-affected populations and nearly 95% of infrastructure losses worldwide.

Actions Proposed:

• Strengthening leadership and cooperation on global water security to address climate-induced challenges such as water scarcity, hazards, and pollution.
• Taking integrated and intersectoral approaches to water-related climate adaptation and mitigation.
• Incorporating perspectives of vulnerable groups, including Indigenous Peoples, women, children, and persons with disabilities, in climate and water-related policies.

 

Drought has left Virginia

Last Thursday, the Interstate Commission on the Potomac River Basin (ICPRB) and the Metropolitan Washington Council of Governments’ (COG) Drought Coordination Committee lifted the Drought Watch declaration that had been in place since July 2024

“The recommendation of ICPRB was based on several factors,” explains Michael Nardolilli, the executive director of ICPRB, “The U.S. Drought Monitor map shows zero signs of drought in the Potomac River watershed. The recent rains have nearly erased the basin’s cumulative precipitation deficits. While spring rains are generally ineffective for groundwater recharge, given the amount of rain the basin has received over the past couple months, staff at ICPRB are optimistic that there will be continued improvement in groundwater levels.”

 As of the end of last week, there were no areas of drought in Virginia, the heavy rains in May and the wet June managed to wipe out the drought. Though water year to date (October 1- June 20^th ) my rain station monitor was still just over 5 inches of rain below average. We’ll see how the rest of the summer plays out, but for the moment we are good.  Due to recent rains across the entire Commonwealth, the Virginia Department of Environmental Quality (DEQ) has lifted drought advisories for 12 counties in Virginia, and no counties remain under a drought watch or warning advisory. The last time the Commonwealth was not under any drought advisory was June 20, 2024.

from DEQ

The past two weeks have brought plentiful rainfall over the entirety of the Commonwealth, further improving soil moisture and surface water flows within all major watersheds. Improvements in groundwater levels have also been observed throughout the State but below normal levels noted in only a few wells in the Eastern Shore, Northern Virginia, Roanoke and Shenandoah Regions. We need much more rainfall to bring groundwater levels to normal levels. All major water supply reservoirs are at normal levels.

A wet summer would be a good thing. NOAA’s outlook for the 2025 Atlantic hurricane season, which goes from June 1 to November 30, predicts a 30% chance of a near-normal season, a 60% chance of an above-normal season, and a 10% chance of a below-normal season. So the outlook is for water. Hopefully enough and not too much. Unfortunately, rain never falls in gentle steady averages.

The NOAA is forecasting a range of 13 to 19 total named storms (winds of 39 mph or higher). Of those, 6-10 are forecast to become hurricanes (winds of 74 mph or higher), including 3-5 major hurricanes (category 3, 4 or 5; with winds of 111 mph or higher). 

from NOAA

Climate Models Biases and Limits

 Regional climate signals pose new challenges for climate science | News | Physical Sciences Division | The University of Chicago

Microsoft Word - _09-09-13_ Chapter 1 Models.doc

Local predictions of climate change are hazy. But cities need answers fast | Science | AAAS

The article below is excerpted from the University of Chicago press release and the other published articles linked above. 

Climate models are good at the big picture of global warming, but at a regional level they have blind spots. Climate science has correctly predicted many aspects of the climate system and its response to increased atmospheric carbon dioxide concentrations in the past one hundred and seventy-five years. However, those are very broad strokes. Recently, discrepancies between the real world observations and expectations for regional climate changes impacts have emerged and been acknowledged.  With decades of observations in hand, researchers can now identify local climate trends the models failed to predict.

• Some of the published examples to emerge are:
• Because warmer air can hold more moisture, models predict rising humidity in many arid places. But such changes have not occurred in the U.S. Southwest and elsewhere.
• Models fail to capture rising rainfall in South America and Australia and drying in East Africa, Europe, and northern India.
• Unforeseen shifts in wind patterns have led to more stalled, sunny weather over Greenland.
• Unforeseen shifts in wind patterns have also strengthened the jet stream over the Atlantic Ocean and intensified storm tracks over the Southern Ocean.
• Europe has faced unexpected summer heat, whereas the Southern Ocean; eastern, tropical Pacific Ocean; and U.S. Midwest have stayed cooler than predicted.

As scientific fields evolve, dominant paradigms emerge and is considered the settled science. When discrepancies or anomalies also arise, they are at first seen as a denial of the “scientific truth,” and often the current framework tries to explain away the anomalies.  However as the anomalies begin to accumulate, the dominant paradigm can be called into question creating a fracture in the scientific consensus. For example, at the beginning of the 20th century, classical physics underwent such a fracture that resulted in the development of quantum physics to understand and explain the anomalies. We may have arrived at such a point in climate science.

According to an analysis by University of Chicago Professor of Geophysical Sciences Tiffany Shaw and Max Planck Institute for Meteorology (MPI-M) Director Bjorn Stevens, published in Nature this month, such a process may be underway in climate science. What the authors describe as the dominant paradigm or “standard approach” of climate science has been developed over the past 60 years by applying fundamental laws of physics to the climate system under the assumption that small-scale processes are determined by statistical averages dependent on large scales (parameterization). This has allowed researchers to uncover the relatively simple physics governing the behavior of the complex climate system.

“The standard approach has been extremely successful in explaining general features of the climate system and certain aspects of its response to increased carbon dioxide concentrations,” said Tiffany Shaw. For example, it does an excellent job of describing and explaining the vertical structure of the atmosphere and some aspects of the spatial pattern of warming of the Earth due to an increase of atmospheric carbon dioxide.

However, Drs. Shaw and Stevens point out that the accumulating discrepancies are exposing knowledge gaps related to assumptions about how large- and small-scale processes and climate system components connect and interact. If these regional discrepancies persist, climate scientists might have to revisit the dominant paradigm.

Given the acknowledged limitations of the current climate models, many climate scientists eschew regional projections. While it may be in foolish to try to use the current regime to make hyperlocal projections, nonetheless there are groups attempting just that because there is little hope at this point of preventing the climate from changing in response to the still ever increasing concentration of carbon dioxide in the atmosphere; our only hope is to know where and how to use our resources to survive.

Chesapeake Bay Watershed Report

Last week the University of Maryland Center for Environmental Science released their 19th annual report card on the health of Chesapeake Bay. Overall, the Chesapeake Bay Watershed scored a C+ (57%). This score is made up of indicators in three categories: Ecology, Society, and Economy. The highest‑scoring category was Ecology, with a B‑ (64%). I’m ignoring the Society and Economy categories because they are not related to the Chesapeake Bay Blue Waters Plan.

There are seven indicators that make up the Bay Health Index for the Chesapeake Bay Report Card; total phosphorus, total nitrogen, dissolved oxygen, water clarity, benthic community, chlorophyll and aquatic grasses. Each indicator is compared to scientifically derived thresholds or goals and scored to determine the overall grade.

from UMCES

Total phosphorus, total nitrogen, dissolved oxygen, and aquatic grasses show positive improvements. These improvements are encouraging for water quality and have positive impacts on the ecosystem. Water clarity and chlorophyll have significantly declining trends. Benthic community shows no significant change in health over time.

Overall Chesapeake Bay Health Scores have been variable in the past and bounced around a bit. However, since 2015, Chesapeake Bay Health Scores have consistently been in the C range (50-57). The ecological score seems to be drifting upward. Improving trends can be attributed to management and restoration efforts at the regional, state, and local levels. Regionally, wastewater treatment plants have been upgraded, and programs have been put in place to reduce nutrient and sediment input. State efforts include seagrass restoration, oyster plantings, and watershed planning. Communities play a huge role in improving Bay health through local resource management activities.

2025 marks a critical deadline for Chesapeake Bay restoration goals, The Chesapeake Bay Clean Water Blueprint established in 2010 as the Watershed Implementation Plan, aims to restore the Bay's water quality by reducing pollution, primarily nitrogen, phosphorus, and sediment. The Blueprint calls for all Bay jurisdictions to have in place, by 2025, the practices and policies necessary to meet the Bay's pollution reduction goals.

Over the years there have been changes in the methods of evaluation. Going forward, University of Maryland Center for Environmental Science plans to improve the Chesapeake Bay Report Card over the next several years. This process will incorporate new indicators for Chesapeake Bay health including some indicators of watershed health.

The new indicators are planned to reflect goals for sustainable fisheries, healthy watersheds, and engaged communities outlined in the Chesapeake Bay Agreement. The planned stakeholder-centered approach will hopefully allow for a stronger report card and improved understanding of how ecosystem health interacts with other factors throughout the Chesapeake watershed.

Year Six of Woodland Restoration

We are just completing the sixth year of my woodland restoration project. My house sits on a bit over 10 acres, about three of them lawn and ornamental gardens. The remaining seven acres is woodland, and much of the woodland is part of the “resource protected area,” RPA, of the Chesapeake Bay. 

When we first moved here, we did not worry about dead trees, as it was all part of the natural process of renewal. A healthy forest has living trees functioning as part of a balanced and self-replacing ecosystem that is a complex mix of trees, understory shrubs and groundcover. In a healthy woodland the process of natural succession occurs over time. Small saplings develop and will become the next generation of trees as the older ones die out. Benign neglect had been my rule for managing the RPA that protects the stream.  

However, about a decade ago, I noticed that something had gone seriously wrong with the forest. The woodland was being destroyed by invasive insects, invasive vines and an explosion of deer and wildlife consuming the native understory. Deer prefer to eat native plants. They devour the saplings of the native trees, but pretty much leave the autumn olive and other invasives alone. When a gap appears in the canopy, there aren’t young trees in the understory waiting for their chance to grow in the sun, but rather, invasive vines and shrubs waiting to take over the landscape.

The woodlands are necessary for a functioning ecology. RPA’s in the Chesapeake Bay Preservation Act are vegetated areas along water bodies, such as lakes, streams, rivers, marshes or shoreline. RPAs are the last line of defense for the protection of water quality in our rapidly developing region. These buffers stabilize stream banks, filter pollutants, reduce the volume of stormwater runoff and provide critical habitat for aquatic species and wildlife. Trees and shrubs in riparian buffers reduce erosion, improve air quality, and provide shade in the summer, windbreaks in the winter and even store carbon.

About a decade ago the number of dead and dying trees had increased dramatically due to the emerald ash borer and it became obvious that the invasive vines, autumn olive and Japanese honeysuckle were choking out the natural renewal process blocking the sun from the trees. So, with guidance from the Virginia Forest Service and the Chesapeake Bay Act guidelines I began a project to restore my woodlands.   

I called the Virginia Department of Forestry to ask for advice. I did not know at the time that Prince William County had its very own forester who can provide information. A Forester came out and inspected the woodland and made some recommendations.  He felt that with removal of the invasive vines and the hanging dead trees the wood might begin to renew itself without requiring any planting. He put his recommendations in a report for me to submit to Clay Morris, Natural Resources Section Chief, Environmental Services Division of Prince William County Public Works to approve the work in the RPA. Though the RPA covers just 2/3 of the woodland, I am treating all the wooded area in the same way. My proposal to Prince William County was strictly by the book in what is allowed in an RPA.

After interviewing several companies, I hired Wetland Studies and Solutions (WSS) to perform and manage the woodland.  I barely knew where to start, the invasive vines blocked entrance to the woods. WSS guided the process and estimated that the project would take 7-10 years to accomplish my goals within my annual budget. They focused our efforts on removing invasive plants and dead or dying trees and tree limbs from the woods and treated the vines that were strangling the native vegetation and causing the most damage. WSSI created strategic brush piles for native animal habitat with the removed dead trees. WSS made sure that all work was conducted in compliance with Prince William County’s Chesapeake Bay Ordinance and all regulations.

WSS created a path from the garden into the RPA using an old deer trail. The trail was overgrown by invasive species and vines, so WSSI’s ecosystems team used foliar herbicide to reestablish the trail and clear out the non-native species. Asiatic bittersweet, autumn olive, and multiflora rose stumps are all spot treated with an herbicide every spring to ensure they would not resprout. Each winter WSSI staff do invasive removals. When the growing season really got underway last week they applied herbicides to the multiflora rose (Rosa multiflora), roundleaf bittersweet (Celastrus orbiculatus), and autumn olive (Elaeagnus umbellata) regrowth. They also sprayed the herbicide to the border of the yard, within the previously foliar treated riparian area, in areas that were cut this past winter, and treated all vegetation on the RPA trail.

It is slow work and and expensive. Every winter a small crew hand cuts the invasive vines and then comes back in the spring to paint the cut stems of the invasive vines with herbicide. We have now just completed the sixth year of  my RPA and forest restoration project and the woodland feels welcoming though I have finally realized that this is probably a perpetual project. I use "we" very loosely, Wetland Studies and Solutions is doing all of the actual labor, my husband and I are simply paying them and directing the sections to be done.  Here are a few pictures of the progress. 

Our trail to the river. Multiflora rose saplings have been greatly reduced in this area.

Area where WSS cut  multiflora rose this past winter. 

Coralberry is now prevalent in the garden boarder.

Now when you walk along the trail you see native species throughout the woods and included wild ginger (Asarum canadense L.)¸ jewelweed (Impatiens capensis), bottlebrush grass (Elymus hystrix), Jack-in-the-pulpit (Arisaema triphyllum), American hog-peanut (Amphicarpaea bracteata), pawpaws (Asimina triloba), and coralberry (Symphoricarpos).  Native tree saplings were present throughout the site and we saw box turtles, deer, turkeys and other native wildlife. The woodland path is a pleasant walk to the river. 

Data Centers Challenge Our Region’s Sustainability

Alissa, H., Nick, T., Raniwala, A. et al. Using life cycle assessment to drive innovation for sustainable cool clouds. Nature 641, 331–338 (2025). https://doi.org/10.1038/s41586-025-08832-3

Using life cycle assessment to drive innovation for sustainable cool clouds | Nature

The article below is excerpted from the article cite above where Microsoft performed a life cycle assessed on the carbon footprint of data centers.

Since 2010, global internet traffic has increased more than 15-fold, with the sharpest jump occurring in the past few years. The increasing demand for cloud apps, machine learning, augmented reality, autonomous vehicles, artificial intelligence (AI) and other applications is driving the growth in data center traffic. This growth has sustainability challenges.

Generally, data centers consume 10–50 times more energy per square foot than typical commercial office buildings. In 2020 data centers accounted for approximately 1.5% (about 300 TWh) of global electricity demand. This percentage is expected to increase with the rapid growth in data centers. It should be noted that energy efficiency improvements in data center energy use has reduced energy intensity of global data centers has 20% annually since 2010, but this rate of improvement has been slowing and the rate of acceleration of data center construction and size has accelerated.

The information technology (IT) industry has benefited from efficiencies following Moore’s law (the number of transistors on a chip and the resulting processing power doubling every 2 years) and Dennard scaling (doubling the transistors per unit area for each new generation of semiconductors without altering their power dissipation). This dynamic has changed and resulted in a slowing of Moore’s law and Dennard scaling in the last few years. Now, improving chip performance requires more power and generates more heat.

Microsoft has committed to being carbon-negative and water-positive by 2030. As part of achieving this commitment, Microsoft has used life cycle assessment (LCA) to systematically analyze the potential environmental impacts of data centers (GHG emissions, energy and water consumption) and enable sustainability by design. When applied to the wider data center industry, LCAs show that traditional cooling technologies can make up 40% of the total energy demand of the data center.

Data center cooling systems, including required equipment, have been extensively reviewed. However, the Microsoft team reports that this is the first public LCA comparing the GHG emissions, energy demand and blue water consumption of air cooling, cold plate and immersion cooling completed by a hyperscale cloud provider. Air cooling uses by far the most electricity. Although hydrocarbon oils used in one-phase immersion cooling are recognized for their dielectric properties, low toxicity and low fluid loss, their high flammability is a safety concern, and their high viscosity creates pumping difficulties. Two-phase immersion can support very high tank power densities (+500 kW) but uses polyfluoroalkyl substances (PFAS) that have been under legislative scrutiny and may be banned entirely. PFAS and requires complex tanks and containment to control the forever chemicals. It would be insane to allow millions of square feet of tanks containing flammable or PFAS solutions all over northern Virginia to save 1-2% of CO2 emissions over direct contact cooling. The solution used for direct contact is flammable and needs to be properly handled and stored.

from Alissa, H., Nick, T., Raniwala, A. et al.air cooling above and plate cooling below

from Alissa, H., Nick, T., Raniwala, A. et al. both one phase and two phase immersion require tanks to submerge the processors

Nonetheless all cooling methods require air-handling units (AHUs) with evaporative cooling to provide cooling for the server rooms. Apart from AHUs for the server rooms, the cold-plate design and the immersion-cooled model uses fluid coolers to reject server heat by a heat transfer fluid. When outside temperatures is below 95 degrees Fahrenheit the fluid coolers can provide sufficient cooling without water (dry operation). However, above 95 degrees additional cooling capacity is required in which case adiabatic cooling (wet operation) is used. Adiabatic systems pre-cool warm outdoor air with water taking advantage of the temperature decrease when water changes phases from liquid to vapor. The bottom line is in Virginia, adiabatic cooling will remain part of the solution to keep data centers operating. Data centers will continue to consume water.  

Alimatou Seck, Senior Water Resources Scientist of the ICPRB found that data centers currently consume about 2% of the water used from the river basin rising to about 8% in the summer when adiabatic cooling is necessary. If the industry continues to grow at an unconstrained pace using standard cooling technologies, it has been widely reported that Dr. Seck  projected that number could surpass 33% by 2050, using 200 million gallons of Potomac water per day. This assumes that the cooling technologies remain the same mix as they are now. That assumption is very unlikely given the information in above study.

Alissa, H., Nick, T., Raniwala, A. et al. Schematic of liquid cooling

Overuse of Groundwater in CRB Threatens Sustainability

 

Declining Freshwater Availability in the Colorado River Basin Threatens Sustainability of Its Critical Groundwater Supplies Karem Abdelmohsen, James S. Famiglietti, Yufei Zoe Ao, Behshad Mohajer, Hrishikesh A. Chandanpurkar; First published: 27 May 2025  https://doi.org/10.1029/2025GL115593

The article below is excerpted from the research cited above and the LinkedIn post of James S. Famiglietti.

From Famiglietti et al

Climate change is causing more frequent and intense droughts around the world, including in the Colorado River Basin, which supplies water to seven U.S. states and Mexico, and is facing severe water shortages. The researchers used data from the NASA GRACE and GRACE Follow-On (GRACE-FO) satellite missions, land surface models, and in situ data over the past two decades to examine and document the groundwater depletion in the Colorado River Basin.

As Colorado River streamflow diminishes, the reliability of surface water resources has become increasingly threatened. Over the past century, the river's flow has decreased by approximately 20%, and climate models predict further reductions of up to 30% by the mid-21^st century due to rising temperatures and reduced snowpack in the Rocky Mountains, which feed the river.

Although all seven states have equal priority under the 1922 Colorado River Compact, under the 2007 Interim Guidelines, Arizona faces the largest mandatory reductions in water deliveries when Lake Mead reaches critically low elevations (Bureau of Reclamation, 2020). Arizona represents a substantial portion of the Lower Colorado Basin in terms of land area and water use, and faces potentially losing up to 512 million cubic meters (approximately 415,000 acre-feet) of its annual allotment if water levels in Lake Mead continue to decline (Bureau of Reclamation, 2020). 

The decline of the river poses a severe threat to both agricultural and municipal water supplies, which are heavily reliant on the river. Approximately 80% of the Colorado River Basin's water is used for irrigation (which supports a $1.4 billion agricultural industry in Arizona alone.

from Famiglietti et al (both charts)

declining groundwater storage in Upper and Lower Colorado River Basin

The declining river flows and reduced allotments places immense pressure on the region's groundwater resources which have been used to make up the difference. As surface water becomes less dependable, the demand for groundwater is projected to rise significantly. In the Lower Colorado River Basin, groundwater already accounts for approximately 40% of the total water supply and is expected to increase. However, the lack of robust protection and management strategies for groundwater in the Colorado River Basin puts these resources at significant risk of over-exploitation. Groundwater is a crucial buffer as water supply, but it is rapidly disappearing due to excessive extraction on one hand and insufficient recharge and management on the other. 

This research highlights the importance of understanding the factors affecting groundwater levels to create sustainable water management strategies that can help secure water resources for the region's future. The research found that:

• The Colorado River basin has lost 27.8 MAF of groundwater since 2003. This is roughly the equivalent of the volume of Lake Mead, and it represents two-thirds of the losses of total water storage.
• Groundwater losses in the Colorado River basin were 2.4 times greater than the losses of surface water in Lakes Powell, Mead, and other smaller reservoirs (11.67 MAF), yet groundwater receives scant little policy and management attention relative to the reservoir levels and the flows of the Colorado River.  
• Most of the losses occurred in the dryer, Lower Colorado river basin, which lost 20.7 MAF of groundwater since 2003, and accounts for 71% of the total water loss (29.2 MAF) in the lower basin.
• Total water storage losses in the Colorado River basin have accelerated by a factor of 3 in the last 10 years compared to the previous decade.

The work suggests that, in states like Arizona, where only 18% of the area of the state has groundwater management, that expansion of groundwater management across the entire state is a critical step towards preserving this precious resource for future generations and for the long-term economic vitality in the region.

PWCA Working for Sustainable Water

 In a briefing to the Board of County Supervisors in December 2024, the Occoquan Watershed Laboratory Director, Dr. Stanley Grant made it clear that emerging water quality issues are a result of the “built” environment. As we continue to develop the Occoquan Watershed we endanger the sustainability of the water supply for up to 1 million people in northern Virginia. When population  density increases, the impervious surfaces in a watershed increase. However, the increase is not linear, once the population density is reaches 100 people per square mile, the rate of increase in impervious surfaces takes off.

from S. Grant presentation

The decisions made in the update to the comprehensive plan two years ago of expanding the development areas of the county, will only serve to increase the rate of degradation of the watershed enabling  rapid sprawl development. The source water to the Occoquan Reservoir is supplied by:

• Bull Run Watershed 25%
• Occoquan River Watershed 48%
• Groundwater and other watershed 20%
• UOSA Reclamation Facility 6%

Prince William County, the Prince William Conservation Alliance (PWCA) is countering this trend with a new initiative designed to guide the county’s growth in fiscally responsible ways that protect natural resources, enhance community character, support a thriving economy. To the typical elements of “Smart Growth” they have added sustainable water since Prince William at 40% of the Occoquan Watershed remains an essential element in sustainable water for the region.

The section below is from “The 3Rs of Reaching Our Potential—Reimagine, Reinvest, Redevelop” a strategic framework to help residents, county staff, the Board of County Supervisors and other decision makers envision a smarter, more sustainable path forward for growth and development.

 

“Drinking Water - The Missing Piece to Our Smart Growth Puzzle

Smart growth often emphasizes walkable communities, protecting farmland, and preserving natural resources. But one essential resource is frequently overlooked: our drinking water.

Clean, reliable, and affordable water is foundational to any community. In Prince William County, this means protecting the Occoquan Reservoir—our region’s primary drinking water source—and preserving groundwater, which supplies 15% of county residents.

Land use decisions directly impact water quality. As we develop the watershed, increasing impervious surfaces like roads and rooftops reduces groundwater recharge and sends more polluted runoff into our streams.

Watershed management and protection must be at the core of smart growth to truly plan for a resilient future. That means concentrating development where infrastructure already exists and avoiding further expansion into the forested and rural areas that protect and relieve development pressure on our drinking water watershed. We especially want to focus protections on reservoir-adjacent lands, which is what ORPA is designed to do, and tributary headwaters, which the AAOD supports.

Data published by the Occoquan Watershed Monitoring Lab suggests that emerging water quality challenges, namely salt and PFAS have smart growth elements to their management strategy i.e. concentrate population growth near transit hubs and limit expanding development further on rural or forested lands in the Occoquan Reservoir Watershed can prevent further degradation of the watershed and the source waters for the Occoquan Reservoir.”

As you can see above, thought UOSA only supplies 6% of the water to the Occoquan Reservoir, it supplies 20% of the salt only half of which comes from households and 39% of the PFOA ( with unknown portion from households). Industrial users are discharging their wastewater to UOSA and that is extremely problematic.