San José State University
Thayer Watkins
Silicon Valley
& Tornado Alley

The Role of Anthropogenic Water Vapor in Earth's Climate

This is a review of the positions taken by various organization concerned with climate about the role of water vapor in the Earth's climate system. Water in its various forms is overwhelmingly the most important substance in climate system. Water vapor is a greenhouse gas, even more effective at absorbing the thermal radiation from the Earth's surface than carbon dioxide. Carbon dioxide has a special role simply because its absorption spectra differs to a degree from that of water vapor.

There has been an attempt on the part of computer modelers to deny the role of water vapor. Yes, they include the positive feedback effect but that is not the only way water vapor is involved in the climate system. In addition to the induced changes in water vapor content there are three other< direct, sources of anthropogenic water vapor: 1. Evaporation and transpiration losses from farming and the watering of lawns and landscapes 2. Evaporation losses from canals and artificial lakes including the reservoirs behind dams. 3. The burning of petroleum and natural gas. Losses from water pumped up from aquifers, fossil water, is especially signficant as an anthropogenic source of increase water vapor in the air.

Consider this record of the water vapor content in the atmosphere above Boulder, Colorado.

Does anyone seriously believe that the these fluctuations in the water vapor content are due to temperature fluctuations or that the trend is due to global warming at a rate of 0.7°C per century. Those who want to ignore the direct water vapor role in climate claim falsely that water uses such as irrigation affect the humidity only in the immediate vicinities of the fields. The graph on the left is for the stratosphere (20-22 km) but the graph on the right shows the trend rates for all levels of the atmosphere. The trend is more radid at the lower altitudes.

The far more reasonable explanation of the trend is the amount of water being put into the atmosphere from farming and urban landscapes in the area.

There is an attempt to justify the neglect of the anthropogenic water vapor by assuming that the atmosphere is at saturation at all places, all levels and all times. Under these conditions the addition of more water vapor from burning fuel only results in water condensing elsewhere. Saturation conditions may prevail at the sea surface but not elsewhere. Would anyone seriously believe that saturation conditons prevail in the land areas such as the western United States? However even if the assumption were valid the effect of condensed water in the atmosphere would be as important as the effect of increased water vapor. Anthropogenic cloudiness is just as important as anthropogenic water vapor, so the saturation argument in addition to being empirically wrong is additionally theoretically invalid.

Now for a review of the comments of various sources on the role of water vapor in the climate system.

National Oceanic and Atmospheric Administration (NOAA)

The comments by NOAA concerning the role of water vapor in global warming are generally judicious and exemplary. Unfortunately they are marred by the presumption that water vapor only enters into the climate system as a feedback effect.

Water Vapor is the most abundant greenhouse gas in the atmosphere […] However, changes in its conentration is also considered to be a result of climate feedbacks related to the warming of the atmosphere rather than a direct result of industrialization. The feedback loop in which water is involved is critically important to projecting future climate change, but as yet is still fairly poorly measured and understood.

As the temperature of the atmosphere rises, more water is evaporated from ground storage (rivers, oceans, reservoirs, soil). Because the air is warmer, the relative humidity can be higher (in essence, the air is able to 'hold' more water when its warmer), leading to more water vapor in the atmosphere. As a greenhouse gas, the higher concentration of water vapor is then able to absorb more thermal IR energy radiated from the Earth, thus further warming the atmosphere. The warmer atmosphere can then hold more water vapor and so on and so on. This is referred to as a 'positive feedback loop'. However, huge scientific uncertainty exists in defining the extent and importance of this feedback loop. As water vapor increases in the atmosphere, more of it will eventually also condense into clouds, which are more able to reflect incoming solar radiation (thus allowing less energy to reach the Earth's surface and heat it up). The future monitoring of atmospheric processes involving water vapor will be critical to fully understand the feedbacks in the climate system leading to global climate change. As yet, though the basics of the hydrological cycle are fairly well understood, we have very little comprehension of the complexity of the feedback loops. Also, while we have good atmospheric measurements of other key greenhouse gases such as carbon dioxide and methane, we have poor measurements of global water vapor, so it is not certain by how much atmospheric concentrations have risen in recent decades or centuries, though satellite measurements, combined with balloon data and some in-situ ground measurements indicate generally positive trends in global water vapor.

The positive trends mentioned above are more likely due to factors such as industrialization, irrigations, urban landscaping and so forth rather than temperature. Here is the temperature records for the U.S.

There has been virtually no increase and what increase there has been was in the winter and spring. However the year-to-year fluctuations indicate that even the winter and spring trends are not significantly different from zero. Therefore if there has been an increaseing trend in water vapor content it has not been induced by temperature increases.

The Wikipedia Article on Greenhouse Gas

This article is virtually useless on the topic because it focuses on irrelevancies such as the structure of models.

Water vapor is a naturally occurring greenhouse gas and accounts for the largest percentage of the greenhouse effect, between 36% and 90%. Water vapor concentrations fluctuate regionally, but human activity does not directly affect water vapor concentrations except at local scales (for example, near irrigated fields).

The assertion that human activity does not directly affect water vapor concentrations except at local scales is merely an assertion not backed by any empirical suport. Additionally the assertion is logically defective in that the effect on the total is the sum of the local effects so necessarily there will be some effect. Furthermore what does the author thinks happens to the water vapor from irrigated fields? Does it fall back into those fields. There are whole rivers of water which flow into the sky. For example, the Colorado River which once flowed into the Gulf of California now disappears before it reaches the Gulf.

The article goes on to say

Current state-of-the-art climate models include fully interactive clouds. They show that an increase in atmospheric temperature caused by the greenhouse effect due to anthropogenic gases will in turn lead to an increase in the water vapor content of the troposphere, with approximately constant relative humidity. The increased water vapor in turn leads to an increase in the greenhouse effect and thus a further increase in temperature; the increase in temperature leads to still further increase in atmospheric water vapor; and the feedback cycle continues until equilibrium is reached. Thus water vapor acts as a positive feedback to the forcing provided by human-released greenhouse gases such as CO2.

On the empirical question of the role of water vapor it is irrelevant what the climate models do or do not do. And the crucial questions concerning models is whether they are valid and validation is achieved by explaining the past climate data with no more information than is available in projecting the future. The models have not been validated, generally because the test has not been applied but in some cases the test has been applied and the model failed but continues to be used projections. Here is the record for the matter of clouds:

Yes, there may be a positive feedback but that does not mean there cannot be a direct effect. The fact that climate models do not include the direct effect states something about the models, not something about the world.

The Magnitude of the Direct (Exogeneous)
Flow of Water Vapor into the Atmosphere

The magnitude of the water that goes into the sky from irrigation and landscaping is enormous. The Colorado River which used to flow into the Gulf of California but is now so diminished that it disappears into the desert carries more gallons of water than all the gallons of gasoline consumed throughout the U.S. An average annual flow rate for the Colorado River upstream is about 10,000 cubic feet per second. That translates into 864 million cubic feet per day, which is about 6.5 billion gallons per day. The total gasoline usage in the U.S. amounts to 378 million gallons per day, only about 6 percent of the disappearance of water from the Colorado.

The Burning of Gasoline

The burning of gasoline generates water vapor as well as carbon dioxide. The chemical reaction formula for burning octane is:

2C8H18 + 25O2 → 16CO2 + 18H2O

So for every molecule of carbon dioxide produced by burning gasoline there is about one molecule of water produced. It is the number of molecules produced which is relevant for the greenhouse effect, not the weight. A gallon of gasoline weighs 6.073 pounds or 2,754.7 grams. One mole of octane is equal to 114 grams so one gallon of gasoline is equal to 24.164 moles of octane. From the chemical equation for the burning of octane we know that 2 moles of octane produces 18 moles of water and 16 moles of carbon dioxide. Therefore one gallon of gasoline produces about 217.5 moles of water. Each mole of water has a weight of 18 grams. So the burning of one gallon of gasoline produces 3,914.6 grams of water. This is equal to 8.6 pounds of water, which has a volume of 1.033 gallons. Thus the water vapor put into the atmosphere by burning gasoline is equal to 390 millions of gallons of water per day. There is of course a near equivalent amount of carbon dioxide put into the air but it is not usually expressed in terms of the equivalent number of gallons of water, but that is the proper metric for a comparison here. Taking into account the relative efficiency of H2O compared to CO2 the amount of CO2 put into the atmosphere by burning 378 million gallons of gasoline per day is equivalent to turning 232 million gallons of water into water vapor.

The Burning of Natural Gas

A similar accounting occurs for the burning of natural gas (methane).

CH4 + 2O2 → CO2 + 2H2O

For methane there are two molecules of water produced for every molecule of carbon dioxide. It is only for the burning of coal that no water molecules are created. There was about 22 trillion cubic feet of natural gas burned in the United States in 2006. A mole of an ideal gas occupies 22.5 cubic meters at 25°C. This volume is equal to 794.6 cubic feet. Therefore the 22 trillion cubic feet of natural gas is equal to 27.7 billion moles of methane. The chemical reaction formula tell us that each mole of methane produces two moles of water vapor when it is burned. Therefore the water vapor produced in a year from burning natural gas is 55.4 billion moles or 996.8 billion grams. One gallon of water weighs 3780 grams so the water vapor produced in a year from burning natural gas is 263.4 million gallons of water. This is per year; per day the figure is 722.4 thousand gallons. This is insignificant compared with the amount generated by burning gasoline.

The Burning of Fuel Oil

The total distillate fuel oil used in the U.S. in 2007 was 64.3 billion gallons, which amounts to 176.2 million gallons per day. Fuel oil weighs 55.6 pounds per cubic foot or 7.43 pounds per gallon or 3370.25 grams per gallon. The molecular weight of fuel oil is roughly 900 grams so the one gallon of fuel oil contains 3.74 moles of molecules 64 carbon units long. The burning of the fuel oil will produce 64 moles of carbon dioxide and 64 moles of water. The gallon of fuel will produce 240 moles of water which is 4,314 grams. This is 4.3 liters of water is 1.14 gallons. Thus the burning of 176.2 million gallons of fuel oil per day produces the water vapor in 200.9 million gallons of water.

Withdrawals from Acquifers

There is also the water from the aquifers such as the Ogallala in the Midwest. Not all of the water pumped from the aquifers goes into the sky but a significant portions does. The Ogallala aquifer has had 312 cubic kilometers or 253 million acre-feet of water pumped out as of 2005. This amounts to 11 trillion cubic feet or 82 trillion gallons. In the year 2000 the withdrawal of water from the Ogallala amounted to 21 million acre-feet. This is 6.8 trillion gallons per year or about 19 billion gallons per day. This is roughly three times the disappearance of the Colorado River.

The U.S. Environmental Protection Agency (EPA) estimates that Americans use 7 billion gallons of water per day for watering lawns and landscapes. There is a comparable though smaller amount used for washing cars, maintaining swimming pools and so forth. The total is thus about 12 billion gallons per day. The EPA notes that about half of this goes into evaporation. The rest gets into the plants and most of this is transpired into the air.


The U.S. Geological Survey (USGS) estimates that in 2000 water withdrawals for irrigation were 137 billion gallons per day. Irrigation withdrawals were 40 percent of total freshwater withdrawals of 342.5 billion gallons per day. Of that 342.5 billion gallons, 134.5 billion gallons were for thermo-electric power plants. Of the non-power plant usage of 208 billion gallons per day . 65 percent was for irrigation.

Fifty eight percent of the irrigation water was from surface water sources. The other 42 percent came from ground water.

About half of the 61.9 million acres irrigated in 2000, 29.4 million acres, were irrigated by flooding the fields. Sprinkler systems were used to irrigate 28.3 million acres and a variety of micro-watering systems were used to water 4.2 million acres. The average application of water to the irrigated land was 26.6 inches of water. The USGS does not provide an estimate of how much of this water, through evaporation and transpiration from plants, goes into the atmosphere but clearly it would be a significant fraction, especially in arid areas where irrigation is necessary. Probably one half is an underestimate but using that as the proportion of irrigation water going into the atmosphere means 68.5 billion gallons per day. That is almost 300 times the effect of the CO2 coming from the burning of 378 million gallons of gasoline per day.

Water for Thermoelectic Power Plants

As stated above, of the 342.5 billion gallons of fresh water withdrawals , 134.5 billion gallons were for thermo-electric power plants. Note that irrigation and power plants use a nearly equal amount of water.


Around the world there is a crisis on river use. When the upstream users draw more water there is less for the downstream users. Where does the water go that is drawn from the rivers by the upstream users. That which does not find its way back to the rivers goes into the sky. Where did 90 percent of the water in the Aral Sea go. Some went into the melons and cotton grown by irrigation, but most of it went into the sky.

The proper unit of measurement for the greenhouse gases is moles. However not all greenhouse gases have the same effectiveness in absorbing thermal radiation. The water molecule is 50 percent more effective than the carbon dioxide molecule. Thus the burning of two moles of octane results in the production of 16 moles of carbon dioxide. The burning of one gallon of gasoline results in the production of 193.3 moles of carbon dioxide. This has the effect of 128.9 moles of water. when this is compared with the 217.5 moles of water actually produced by the burning of a gallon of gasoline we see that 63 percent of the effective greenhouse gas from burning gasoline comes from the water vapor it produces. This does not take into account the resident times for the two gases. Nevertheless the cursory estimates indicate that direct anthropogenic water vapor dwarfs the amount of carbon dioxide going into the atmosphere from burning hydrocarbon fuels.

A water molecule does not stay in the atmosphere as long as a carbon dioxide molecule does. However what matters is the amount that is in the atmosphere at any one time. However one effect of anthropogenic water vapor is an increase in precipitation downwind from the regions adding water vapor to the air. One of the puzzles of U.S. climatology is that there has been no statistically significant increase in temperature yet there has been statistically significant increases in precipitation. It is no puzzle. The water evaporated from farm fields, urban lawns and swimming pools falls as rain and snow to the areas to the east. Some of the rainfall then evaporates and falls again still farther to the east.

The overwhelming conclusion from the above is that there is no reason to neglect the direct anthropogenic increases in the atmosphere on climate. There are water vapor feedback effects induced by both the anthropogenic carbon dioxide and direct anthropogenic water vapor.

HOME PAGE OF applet-magic
HOME PAGE OF Thayer Watkins