United States
            Environmental Protection
            Agency
              Office of Atmospheric
              and Indoor Air
              Programs
   EPA430-R-93-011
   August 1993
vvEPA
Current and Future
Methane Emissions From
Natural Sources
            Report to Congress
                         PROPERTY Of
                           DIVISION
                             OF
                         METEOROLOGY
     Boreal Wetlands
          Permafrost
Arctic Tundra
      Offshore
     Gas Hydrates
       Amazon Floodplain
 Congo Basin

-------
  Current And Future Methane
Emissions From Natural Sources
       Report to Congress
        Editor: Kathleen B. Hogan
   U.S. Environmental Protection Agency
        Office of Air and Radiation
            August 1993

-------
                              TABLE OF CONTENTS
EXECUTIVE SUMMARY  	   ES-1

INTRODUCTION	  1-1
      1.1   Methane as a Greenhouse Gas	  1-1
      1.2   Present, Past and Future Sources of Atmospheric Methane	  1-1
      1.3   Current Emissions from Natural Systems	  1-4
      1.4   Future Emissions from Natural Sources	  1-9
      1.5   Overview of this Report	   1-10

NATURAL WETLANDS	  2-1
      2.1   Background -- Wetlands Classification   	  2-1
            2.1.1  Tropical Wetlands 	  2-2
            2.1.2  Temperate Wetlands  	  2-3
            2.1.3  Northern Wetlands	  2-3
      2.2   Review of Emission Measurements	  2-5
      2.3   Global Emissions  	   2-10
      2.4   Effects of Environmental Variables	   2-13
            2.4.1  Emission Rate  	   2-15
            2.4.2  Wetland Area and Emission Period	   2-22
      2.5   Possible Future Scenarios for Methane Emissions	   2-24
            2.5.1  Predicted Climate Change	   2-24
            2.5.2  Developing Future Estimates of Emissions	   2-25
      2.6   Uncertainty and Further Research Needs	   2-33
            2.6.1  Current Emissions  	   2-34
            2.6.2  Future Emissions  	   2-38
            2.6.3  Methane Uptake in Wetlands	   2-40

NATURAL FOSSIL SOURCES	  3-1
      3.1   Gas  Hydrates	  3-1
            3.1.1  Background	  3-1
            3.1.2  Current Emissions  	  3-5
            3.1.3  Future Emissions  	  3-6
      3.2   Permafrost  	   3-10
            3.2.1  Background	   3-11
            3.2.2  Current Emissions  	   3-12
            3.2.3  Future Emissions  	   3-13
      3.3   Uncertainty / Further Research	   3-13

SUMMARY AND CONCLUSIONS	  4-1

REFERENCES	 RF-1

APPENDICES	   AP-1
      Appendix A: List of UNH Workshop Participants 	   AP-1
      Appendix B: Methane Flux from Tropical Wetlands	  AP-2
      Appendix C:  Methane Flux from Temperate and Subtropical Wetlands	  AP-4
                                                                            Page i

-------
Appendix D:  Methane Flux Measurements from Northern Wetlands	  AP-8
Appendix E:  Methane Flux from Tussock Tundra (Warm Season Only) ....  AP-15
Appendix F: Mean Regional Methane Fluxes (High Latitudes)	  AP-17
Appendix G:  Rates of CH4 Uptake in Natural Ecosystems	  AP-18
                                                                    Page ii

-------
                                 LIST OF EXHIBITS

EXECUTIVE SUMMARY
Exhibit ES-1   Natural Sources of Atmospheric Methane	   ES-2
Exhibit ES-2   Methane Emissions from Natural Wetlands	   ES-3
Exhibit ES-3   Summary of Northern Wetland Emission Scenarios	   ES-7
Exhibit ES-4   Potential  Increases  in  Methane  Emissions  from  Selected  Natural
             Systems due to Climate Change	ES-10

INTRODUCTION
Exhibit 1-1    Sources of Atmospheric Methane	   1-2
Exhibit 1-2    Methane Concentrations and Temperature Changes in the Past
             160,000 Years	   1-3
Exhibit 1 -3    Methane Emission Pathways	   1-5
Exhibit 1 -4    Wetlands Map	   1-6
Exhibit 1-5    Global Wetland Area	   1-6
Exhibit 1-6    Location of Known or Inferred Gas Hydrates	   1-8
Exhibit 1-7    Climate Feedback to Doubled  C02	   1-10

NATURAL WETLANDS
Exhibit 2-1    Measurements  in Tropical Regions Summarized in this Report	   2-6
Exhibit 2-2    Average Emission Rates from  Tropical Wetlands   	   2-7
Exhibit 2-3    Average Emission Rates from  Northern Wetlands	   2-10
Exhibit 2-4    Global Tropical  Methane Emissions	   2-11
Exhibit 2-5    Global Temperate Methane Emissions	   2-12
Exhibit 2-6    Global Northern Methane Emissions	   2-12
Exhibit 2-7    Global Annual Methane Emissions from Natural Wetlands	   2-14
Exhibit 2-8    Comparison of  Global Emission Studies	   2-15
Exhibit 2-9    Effects of Environmental Variables on CH4 Emission Rate	   2-17
Exhibit 2-10   Emission Rate vs Soil Temperature in Northern Wetlands  	   2-19
Exhibit 2-11   Regression of Emissions Rate  Data in Exhibit 2-10	   2-20
Exhibit 2-12   Effects of Environmental Variables on Wetland Area and Emission
             Period	   2-23
Exhibit 2-13   Predicted Temperature Change (50-70 N)  	   2-26
Exhibit 2-14   Oak Ridge Workshop Estimates of Increased Northern Ecosystem
             Methane Emissions	   2-30
Exhibit 2-15   Methane Feedback Processes in a Warmer, Drier Climate	   2-32
Exhibit 2-16   Methane Feedback Processes in a Warmer, Wetter Climate	   2-33
Exhibit 2-17   Summary of Northern Wetland Emission Scenarios	   2-34
Exhibit 2-18   Uncertainty in Current Wetland Emission Estimates  	   2-36

NATURAL FOSSIL SOURCES
Exhibit 3-1    Relative Location of Hydrate Types   	   3-2
Exhibit 3-2    Methane Hydrate  Reserves 	   3-3
Exhibit 3-3    Hydrate Stability Zone 	   3-4
Exhibit 3-4    Oceanic Hydrate Scenarios 	   3-9
Exhibit 3-5    Continental Hydrate Scenarios	   3-10
                                                                             Page iii

-------
SUMMARY AND CONCLUSIONS
Exhibit 4-1   Potential Increases in Methane Emissions from Selected Natural
            Systems due to Climate Change	  4-2
                                                                            Page iv

-------
                               ACKNOWLEDGEMENTS
      This report was made possible with the intensive efforts of a number of people. Robert
Harriss, Patrick Crill, and Karen Bartlett of the University of New Hampshire provided the
foundation for the chapter on emissions from wetlands by drafting a state-of-the-art review
of experimental data on methane emissions from wetlands, in addition to organizing an expert
workshop for discussion of important issues affecting future emissions from these systems.

      The participants in the expert workshop were Karen Bartlett, Jeffrey Chanton, Patrick
Crill, Stephen Frolking, Robert Harriss, Harold Hemond, Kathleen Hogan, Michael Keller, Daniel
Lashof,  Elaine Matthews, William Pulliam,  Jeffrey Ross, Nigel Roulet,  Margaret Torn, and
Steven Whalen. They contributed important information and comments in addition to helping
frame the different ways of looking at the problem of estimating future emissions.

      Keith Kvenvolden,  E.G. Nisbet, and  T.E. Osterkamp provided valuable information,
comments and insights for the chapter on emissions from fossil sources.

      Jeffrey Ross drafted several sections of the report, including the chapter on emissions
from fossil sources. He also played the key role of document integration and editing as well
as providing technical and graphical support.

      Many people  provided comments on different versions of the  report.   Particularly
important comments were provided  by Karen Bartlett, Patrick Crill,  Robert Harriss, Daniel
Lashof and researchers (Jack Jones,  David Lewis, and Lee Mulkey) at EPA's Environmental
Research Laboratory in Athens, Georgia.
                                                                               Page v

-------
                               EXECUTIVE SUMMARY

      This report is one in a set of reports requested by Congress in Section 603 of the Clean
Air Act Amendments  of 1990 to  provide information on  a variety  of  domestic and
international methane issues.  This report provides estimates of current and future methane
emissions from natural  sources.

Introduction

      Methane is a "greenhouse gas," meaning that its presence  in the atmosphere affects
the earth's temperature.  As a greenhouse gas, methane is  a large contributor to potential
future warming of the  earth.   Its concentration in the  earth's atmosphere has more than
doubled over the  last two centuries  (after remaining fairly constant for the previous 2,000
years), and continues to rise. Methane is second only to carbon dioxide in its contribution to
potential future warming  of the earth.

      Methane's increasing atmospheric concentration  is largely correlated with increasing
human population.  Human-related activities such as fossil fuel production, transportation,
animal husbandry, rice  cultivation, and waste management release significant quantities of
methane, and all of these activities are expanding with industrialization and population growth.
It is well established that  these sources currently represent about 70 percent of total annual
emissions. Natural sources account for the remaining 30 percent of methane emissions (See
Exhibit ES-1).

      While recent increases in atmospheric methane concentrations are largely attributed
to human activities, variations in methane's atmospheric level over the  previous 150,000
years are  largely attributed  to changes in methane emissions from natural systems, and in
particular wetlands. This  experience, in addition to other recent research, suggests that there
is  potential for the methane emissions from natural sources  to  increase as climate  changes
in the future. The emissions from several of the  natural sources -- in particular, wetlands, gas
hydrates,  and  permafrost --  are strongly governed by environmental  variables  such as
temperature and precipitation.  Therefore, climate change induced by humans  could actually
trigger the release of more greenhouse gases from natural systems.  The implications of this
effect are twofold:

            The rates and magnitude of future climate change would increase.

            The problem of controlling emissions of the greenhouse gases, so as to reduce
             the adverse effects of climate change, would be greatly exacerbated.
       This   report  investigates  current
methane emissions from natural systems by
examining emission data that have become
available in the last several years (and  for
the most part have not been reflected in
This report f o
-------
Exhibit ES-1
Natural Sources of Atmospheric Methane
Source
Wetlands2
Termites
Oceans
Freshwater
Gas Hydrates
Permafrost
Total for Natural Sources
Total Methane Emissions3
Emissions
Estimate
(Tg CH^yr)1
109
20
10
5
5
0
150
505
Range
(Tg CH4/yr)
70-170
10-50
5-20
1-25
0-5
?
100-300
400-610
Source: IPCC 1992.
1 Tg = teragram = 10|12 grams.
2 taken from this report.
3 Crutzen 1991.
IPCC1 assessments). By further investigating recently available scientific information on the
major processes governing methane emissions  from these ecosystems, estimates of the
potential for emissions to increase in the future, as a result of climate change, are developed.
Emphasis is given to the three systems that are most likely to be affected by climate change -
- wetlands, gas hydrates, and permafrost.

Current Methane Emissions from Wetlands, Gas Hydrates, and Permafrost

       Wetlands represent between 4 percent and 8 percent of the earth's land surface and
are currently the primary source of methane emissions from  natural sources.  Methane is
generated in moist, oxygen-depleted, wetland soil by bacteria, as they decompose dead plant
material. Emissions from gas hydrates and permafrost currently represent less than 4 percent
of emissions from natural sources.  However, they hold vast reserves of methane that may
be liberated as temperatures rise.
   1 The Intergovernmental Panel on Climate Change (IPCC) was established in 1988 under the auspices of the
World Meteorological Organization and the United Nations Environment Program.  Among other tasks, the IPCC
was charged with assessing the science underlying the greenhouse effect. One assessment was published in
1990, and an update was published in 1992.  Much of the data presented in this report is included in these
science  assessments, however, substantially more data has become available since their drafting.
                                                                                Page ES-2

-------
      Wetlands

      Several recent studies have estimated that natural wetlands currently emit about 110
Tg of methane per year. These works include Matthews and Fung, 1987 (110 Tg/yr) and
IPCC, 1990 (115 Tg/yr, with an uncertainty range of 100 to 200 Tg/yr).

      In the last four  years, many experimental studies  were completed  which greatly
increase the number of available measurements of methane emissions from wetland systems.
Many of these  studies were performed  in tropical  and northern  areas  where there was
previously very limited information.

      Based on analyses in this report, which include this new data, wetlands worldwide are
estimated to emit 109 Tg of methane annually. This global  estimate  is similar to past
estimates of methane emissions from wetlands. However, a difference from previous studies
is in the contribution to global emissions from each region. Here, 60% of wetland emissions
are estimated to be from tropical  systems, whereas  previously tropical  wetlands were only
estimated to contribute 29% to 54% of  the total.   Systems in the northern latitudes now
represent a smaller portion (35%) than in the  past (31% to 58%),  and  temperate wetlands
continue to represent a  small portion (5%) of the total (see Exhibit  ES-2).
Exhibit ES-2
Methane Emissions from Natural Wetlands
Wetlands
Ecosystem
Tropical
Temperate
Northern
Total
Area
(x10nm2)
19
4.1
88.3
111
Annual
Emissions
(Tg CH4/yr)
Value
66
5
38
109
Range
34-118
not
available
29-52
70-170
Percent of
Total
Emissions
60%
5%
35%
100%
       Methane emissions from wetlands are grouped into major regions as follows:

       Tropical Wetlands. Tropical wetlands (those between 20N and 30S) represent 17%
       of total wetland area and 60% of methane emissions from wetlands.  These relatively
       high emissions are due to the higher temperatures and higher levels of solar radiation
       in the tropics as compared with other regions. Tropical wetlands will actively produce
                                                                           Page ES-3

-------
      and emit methane anytime that there is sufficient precipitation to maintain inundation,
      or flooded conditions. The non-forested swamp system has the highest emissions of
      all tropical systems. This system is characterized by rapid plant growth rates and rapid
      decomposition, and average annual flux rates have been estimated at  85 g/m2.  The
      non-forested swamp system represents 42% of tropical wetland area and 52% of
      emissions.

      Northern  Wetlands. Northern wetlands  (those above 45N) are usually underlaid with
      near-surface permafrost which prevents soil drainage and  thus  creates  flooded
      conditions.  Temperature, or the length of the thaw season, largely determines when
      northern wetlands will actively emit methane, as opposed to tropical wetlands, where
      the emission period is determined by precipitation. The majority of emissions from this
      region come from inundated  boreal (between 45N and 60N) and inundated arctic
      wetlands (above 60N), with a small amount coming from well-drained arctic wetlands
      or tundra. The area and average emission rate from the two inundated wetland regions
      are almost identical, but the annual period of emissions is longer in the  more southern
      boreal region.  Therefore, inundated boreal  wetlands  account for 53% of northern
      emissions, compared with 37% from inundated arctic wetlands. The average annual
      flux rates have been estimated at 35 g/m2. Also included in northern wetlands is well-
      drained arctic tundra.  This expansive area (53% of all wetland area) is only marginally
      inundated and, thus, has a very low average  annual emission rate (3 g/m2).  Well-
      drained tundra accounts for 10% of northern emissions  and  4% of total wetland
      emissions.

      Temperate Wetlands.  Wetlands in  the  intermediate  latitudes (between 20N  and
      45N, and between  30S  and 50S) exhibit characteristics of both northern  and
      tropical wetlands. However, because the temperature and rates of precipitation  and
      solar radiation are lower in  this region, and topographic relief is greater, temperate
      wetlands have lower emission rates than tropical wetlands.  Because soil drainage
      tends to be better in the  temperate zone,  wetlands do not extend over nearly as large
      an area in the temperate region as in the  far north. Temperate wetlands account for
      4% of the total wetland area and 5% of  total wetlands methane emissions.

      Based on the analysis in this report, total wetland emissions may be as low as 70 Tg
and as large as 170 Tg.  This wide range arises primarily from the  large variation  in observed
emission  rates from similar  ecosystems. For example, the emission rate for non-forested
tropical swamps is estimated here to be anywhere  from 52 to  285 mg  CH4/m2/day.  In
general, the emission rate is sensitive to a number  of variables including:

           the temperature, since methane-producing bacteria are generally more active
            as temperature increases;

           the level  of the water table, since the area must  be sufficiently flooded to
             maintain anaerobic conditions; and

            the  plant community, since  the  plants affect the availability of carbon for
             decomposition, in addition to the transport of methane from the anaerobic zone
            to the atmosphere.
                                                                           Page ES-4

-------
      While large uncertainties in emission rates remain, the growing body of experimental
information is beginning to show consistent results for experiments performed  on similar
wetland systems in different areas.

      An attempt is also made to include the uncertainty in the total wetlands area in the
range for total wetland emissions. This is accomplished by comparing two different surveys
of wetland areas.  However, uncertainty in  the period -- the length  of time that flooded
conditions are maintained over the course of a year -- is not explicitly accounted for in the 70
to 170 Tg range. This source of uncertainty is not accounted for because  no acceptable
method  has been found for quantifying the uncertainty in period.

      Gas Hydrates and Permafrost

      Gas hydrates and permafrost are two  systems that currently contribute little, if any,
to the annual  emissions of methane.  However, they both contain substantial reserves of
methane and may contribute methane emissions in  the future if climatic conditions change
significantly.

      Gas Hydrates. Methane can be trapped in gas hydrates, which are dense combinations
      of methane and water molecules located deep under the ground and beneath the ocean
      floor. An immense quantity of methane is trapped in both oceanic and continental gas
      hydrates, with estimates ranging  from millions to billions of teragrams.2  Extensive
      information is known about the  temperature and pressure conditions required to
      maintain the stability of hydrates and keep the methane trapped. Scientists generally
      believe that the stability  conditions have been altered for a  small  portion of the
      hydrates as a  result of sea level rise which has occurred since the last major ice age.
      Calculations show that a relatively small amount of methane - 3 to 5 Tg  per year --
      may be escaping to the atmosphere from this region.

      Permafrost. Permafrost is ground, usually consisting  of soil and ice, that  remains at
      or below O'C throughout  the year for at least two consecutive years.  Research has
      shown that methane is trapped in permafrost in small concentrations. Due to the large
      amount of permafrost that exists on earth, the total amount of methane stored in this
      form  could be quite high -- possibly several thousand teragrams. While it has been
      proven that permafrost is  melting in certain locations, no estimates have been made
      for current emissions from this source.

Future Methane Emissions from Wetlands, Gas Hydrates, and Permafrost

      The emissions of methane from wetlands, gas hydrates, and permafrost are strongly
linked to environmental variables, such as temperature and precipitation.  While substantial
uncertainty remains  about how emissions from these systems will respond to changes in
environmental variables, several  of the relationships are becoming better  understood.  For
example:

      Precipitation.  Inundation,  or sufficient soil moisture, is a prerequisite for the anaerobic
      conditions that allow methane generation in wetlands.  Increased precipitation could
   2 1 teragram (Tg) is equal to 1x1012 grams or 1 million metric tons.



                                                                            Page ES-5

-------
      enlarge the area of land that is inundated and generating methane, as well as raise the
      average rate at which methane is generated on a unit area basis.

      Temperature.  A number of experiments have shown that methane emissions from a
      particular wetland  area  may  increase  exponentially  with  rising temperatures.
      Temperature is also the most important factor affecting the stability of  gas hydrates
      and permafrost.  If temperature rises, these two systems could be destabilized, and
      more methane released.

      While changes in soil  moisture and temperature are probably the most important
variables that  will   determine   future  methane  emissions,  changes in several  other
environmental variables could  play important roles. These include the species and density of
plants, human land use impacts (particularly in tropical wetlands), the depth to  permafrost in
northern wetlands, and sea level rise. However, little work has been performed  to assess the
potential effects of these variables.

      There are a number of predictions  for how the  climate might change over the next
century. These projections are based on an effective doubling of carbon dioxide concentration
in the atmosphere, which could  occur by 2050.  Projections have been made for increasing
temperatures and changing precipitation patterns and may be summarized as follows:

           Global  average  temperature increases of about 1.9  to 5.2C  have been
             estimated  for  doubled carbon  dioxide  conditions  (Mitchell et  al., 1990).
             Regional  and seasonal increases  may  vary.  For  example,  temperatures
             increases of 4 to 8C  in northern latitudes during  winter are expected.
             Tropical regions are expected to warm less, by about 2 to 3C throughout the
             year.

           Precipitation increases globally, which could increase soil moisture in wetland
             regions.  Conversely,  a corresponding  increase in  evaporation due to higher
             temperature may lead to constant or decreased soil moisture.

Although there are still many uncertainties in these projections (and no predictions at all for
some important environmental variables), they provide a useful basis for investigation  of the
effects  that a continuation of current energy use and agricultural practices  may have on
methane emissions from wetlands, gas  hydrates, and permafrost in the future.

      Wetlands

      While much uncertainty remeiins, climate change could cause methane emissions from
wetlands to increase substantially within the next century.  Potential increases in emissions
are supported by two recent scientific workshops (Post 1990; this report) and work by Lashof
(1989). These efforts developed rough quantitative estimates of possible increased emissions
from northern wetlands.  No predictions have  been attempted for tropical wetlands.

      Northern Wetlands.   Methane emissions  from  these  wetlands  could increase
      significantly because northern wetland emissions are believed to be determined largely
      by temperature and because there  is substantial carbon available, in the form of peat
      deposits, which could  be liberated as methane in the future.  Several  scenarios are
                                                                            Page ES-6

-------
                                      Although much  uncertainty  remains,
                                      researchers  estimate   that  global
                                      warming could cause annual methane
                                      emissions from northern wejitsndat to
                                      increase fcy 11 W& T0 &y the mJ of
                                      the century, depending OR climatic
examined in this  report.   In  one
scenario,   temperatures   and
precipitation increase, and wetlands
become   wetter  due  to   more
precipitation.  In a second scenario,
temperatures increase, and wetlands
maintain their current water status as
changes   in   precipitation   and
evaporation offset each other.  In a
third  scenario,  temperatures  and
precipitation   rise,   but  wetlands
become drier because evaporation increases by more than precipitation.

       Given these climate change scenarios, researchers have estimated the potential
magnitude of future  methane emissions.  In all  three scenarios, projected methane
emissions will not decrease significantly and could increase by several fold within the
next century (see Exhibit ES-3).  While it is  likely that the area of northern wetlands
will be altered by future climate change, the predictions made to date do not account
for  any variations in wetland areas.   (A change in  wetland area could increase  or
decrease emissions.) This simplifying assumption is made because of the difficulty in
predicting changes in wetland area.
Exhibit ES-3
Summary of Northern Wetland Emission Scenarios
(values shown are for increased emissions in Tg CH4/yr)
Reference
Lashof 1989
Post 1 9901
UNH Workshop1
Warmer/Drier
~
290
5 -35
Warmer/Wet
17-63
290
35
Warmer/Wetter
-
-
65
1 Expert workshops which emphasized best scientific judgement,
due to limited experimental information in this area.
Tropical Wetlands. To date, predictions have not been made for future emissions from
the tropics.  Predictions for the tropics are even harder to make than are predictions
for the north because emissions have not been shown  to be strongly controlled by
temperature.  Future  tropical wetland emissions will depend on several variables
including  regional  precipitation,  interactions between  precipitation  and  actual
evapotranspiration,  and effects  of human activities.  However, it is not  known how
these variables will change in the future or exactly what their effect will be on methane
emissions.  Also, unlike most northern wetlands, a readily available source of carbon
                                                                      Page ES-7

-------
for increased emissions does not currently exist.  While additional carbon could be
supplied by increased plant activity, the likelihood of this outcome is unknown.
Gas Hydrates
                                     Although  much uncertainty  remains,
                                     scientists estimate that global warming
                                     could cause methane emissions from
                                     gas hydrat&s to rise from about 5 Tg/yr
                                     now to 100 to 1000 T$/yr ove*  the
                                     next few centuries.
      Global warming could jeopardize the
stability of currently stable hydrates, which
contain thousands of teragrams of methane.
While scientists disagree on the exact time
lag   before  climate   change  would
significantly  affect  the   deeply   buried
hydrates -- estimates range from about one
hundred years to a few thousand years --
there is  a  consensus   that  increasing
temperatures  wilt  eventually  destabilize
much of the existing hydrates.

      When warming does reach the hydrates there is a potential for tremendous quantities
of methane  to be released and for some of the methane to escape into the atmosphere.
Several researchers have constructed scenarios of methane hydrate emissions to quantify the
potential risk from this source.  Results of these scenarios are increased emissions of 50 to
300 Tg CH4/yr from continental hydrates and 150 to 640 Tg CH4/yr from oceanic hydrates,
beginning anywhere from  one hundred to several thousand years from now.  Because a
variety of assumptions were used by the different researchers (including assumptions about
future temperature increases),  these scenarios  are "adjusted" in  Chapter 3 to reflect a
common set of assumptions and arrive at composite scenarios for each type of gas hydrate.
The composite  scenarios derived in this report predict emissions of 100  Tg CH4/yr  from
continental hydrates, and 200 Tg CH4/yr from oceanic hydrates, with an uncertainty range
of at least one order of magnitude.  The average prediction of the time  lag before emissions
reach these levels is a few hundred years.

      Permafrost
                                     Global warming smilct cause methane
                                     emissions from p0/msfrosf to reach 60
                                     Tg/yr within the next century*
       There is also potential for significant
quantities of methane to be released from
permafrost to the atmosphere in the future
because large amounts of permafrost could
melt due to rising temperatures in the polar
regions.  Permafrost releases are predicted
to be smaller than hydrate releases, mainly
because there  is much less methane trapped in permafrost than in hydrate form. However,
releases could be as high as 60 Tg  CH4/yr by the end of the next century.

       Summary

       Available information indicates that methane emissions from natural sources could
increase by 5 to 370 Tg/yr by the end of the next century due to projected climate change.
An increase of 5 to 370 Tg/yr is equivalent to about 100 to 7,000 million metric tons of CO2,
                                                                     Page ES-8

-------
or an additional 0.2 to 17 percent of carbon
dioxide beyond current  predictions for the
year 2100.  Furthermore, due to the long
delay before  gas  hydrate emissions  will
increase, natural methane emissions could
increase  by  hundreds  more  teragrams
annually in the centuries to follow the next
one.  The potential increases in emissions
from northern  wetlands, gas hydrates, and
permafrost   for different  scenarios  are
summarized in Exhibit ES-4.

Uncertainties and Needs for Further Work
Climate change; as projected based on
current patterns of energy  use  and
agricultural practices, could trigger the
release ef an additional  S to 37O Tg
H4/yr f ran* natural sources fey H* oad
of the next century.  Farther Increases
in methane  emissions oft the  order of
hundreds of teragrams annually could
occur  over  the next several
years.
       The generation, storage, transport, and release of biogenic methane are highly complex
and variable processes.  Scientists have only recently begun to understand the mechanisms
of these processes and the potential for a natural methane  feedback to climate change.
Significant uncertainty surrounds many of the results presented in this report.  With additional
research in the field of natural methane emissions, this uncertainty can be reduced.

       Wetlands

       An important factor contributing to uncertainty in current emissions estimates is the
wide  variety of  wetland types and  the variability within each type.  Understanding the
magnitude and dynamics of methane emissions at  one site of a certain type, does not
necessarily transfer to other types or even other sites of the  same general "type" that are
geographically remote.  While  several independent studies have arrived at similar global
methane emissions estimates despite the existing uncertainties, more field  research could
further reduce these uncertainties. This research should focus  on  systems not previously
measured, in addition to developing better information on areas of different ecosystem types
and periods of inundation.

       Greater uncertainty exists in the future wetland emission scenarios. Not only are the
relationships between methane emissions and environmental  variables (e.g.,  precipitation,
temperature, actual evapotranspiration, plant community, human impacts, sea level change)
not well known,  but how these  environmental variables will change in the future is also
uncertain.

       The results of the three prediction efforts discussed  here  are  not  intended to be
definitive, only rough ball-park estimates. They are based on the best available methods for
estimation,  which are  crude,  simplified methods.    In some cases the calculations are
reproducible because they are  based on process based methodologies, in other cases the
results are not reproducible because they are based on expert opinion of processes that have
not been modeled. In the opinions of the scientists who developed them, the results are the
reasonable estimates that can be made based on the information available at the present time.
They are intended only  to give  policymakers a rough idea of the order of magnitude of the
change in methane emissions from wetlands that could take place as a result of climate
change.
                                                                             Page ES-9

-------
Exhibit ES-4
Potential Increases in Methane Emissions from Selected Natural Systems due to
Climate Change
System
Tropical Wetlands
Temperate Wetlands
Northern Wetlands
Hydrates (Continental)
Hydrates (Oceanic)
Permafrost
Total
Current Emissions
(Tg CH4/yr)
estimate
66
5
38
5
0
0
114
range
34 - 1 1 8
n/a
29 -52
0-5
0
0
70-175
Range of Predicted Increases
(Tg CH4/yr)
within 100 yrs
O1
O1
5 - 290
O2
O2
0-602
5 -370
after 1 00 yrs
O1
O1
5-290
50 - 3002
1 50 - 6402
0- 602
200 - 1 300
1 In the absence of future predictions for these systems, they are assumed here to
remain constant.
2 The predicted increases for hydrates and permafrost are relatively more uncertain
than the predicted increases for wetlands.
      More field research, especially multi-year flux and environmental variable studies, will
help clarify  how  methane  emissions  are controlled  by factors such  as water level  and
temperature. Natural ecosystem manipulation (artificially altering one variable in an otherwise
unchanged natural ecosystem) and long-term monitoring of "early warning"  or indicator
ecosystems will also improve understanding of wetlands' response to climate change.

      It is essential that process-based wetland models that incorporate the relationship
between emissions and many environmental variables, not just temperature, are developed.
Of course, such models can only be as good as the  environmental data fed into them.
Therefore, it is also important that site specific predictions for the important environmental
variables can be made. As the scope and resolution of general circulation models increases
and  as  process-based wetland models account for more environmental variables,  future
emission scenarios will become more reliable.

      Gas Hydrates and Permafrost

      In general, less is known about the topic of methane emissions from gas hydrates and
permafrost than about wetland emissions. The greatest uncertainty surrounds future potential
emissions from gas hydrates and melting permafrost. This uncertainty arises largely because
gas  hydrates, and especially permafrost, have only recently been  recognized as potential
                                                                           Page ES-10

-------
sources of future methane emissions, and very little research has been performed to examine
this possibility. An extensive sampling program of permafrost and sediment from the hydrate
zone will help resolve some of the uncertainty surrounding methane reserve sizes. Additional
research is also needed to better understand the mechanisms by which temperature changes
are transmitted to the areas of trapped methane, and the processes by which released
methane passes through the water and/or soil column to reach the atmosphere. While this
uncertainty calls into question the timing and magnitude of fossil source emissions, it does
not undermine the conclusion that significant emissions from fossil sources could occur as a
result of climate change.
                                                                           Page ES-11

-------
                                     CHAPTER 1

                                   INTRODUCTION
      This report is one in a set of reports requested by Congress in Section 603 of the Clean
Air Act  Amendments of  1990  to provide information on  a variety  of domestic  and
international methane issues.  This report provides estimates of (1) current methane emissions
from natural sources worldwide,  and (2) how these emissions may change in the future.
Emissions are defined  as methane that escapes  into the atmosphere.  This  report is not
intended to address the significance of methane  in the atmosphere or natural sinks of
methane.

1.1   Methane as a Greenhouse Gas

      Methane (CH4) is one  of a group of atmospheric trace gases that plays an important
role in atmospheric chemistry and in the earth's  energy balance.  Like other "greenhouse
gases,"  methane traps outgoing infrared radiation from the earth's surface and increases the
temperature of the earth.  Methane is also a large contributor to potential future warming of
the earth: its atmospheric concentration has more than doubled over the last two centuries
(after remaining fairly constant for the previous 2,000 years), and its concentration continues
to rise (IPCC 1992). CH4 will contribute about 15 percent of this future greenhouse warming,
second  only to carbon dioxide (Rodhe 1990; IPCC 1992).

1.2   Present, Past and Future Sources of Atmospheric Methane

      Methane's increasing atmospheric concentration is largely correlated with increasing
human populations due to intensified and expanded activities such as the production and use
of fossil fuels,  animal husbandry,  rice cultivation, and waste management. These sources
currently represent about 70 percent of annual methane emissions (see Exhibit 1-1) (Cicerone
and Oremland 1988; IPCC 1990; IPCC  1992).  The remaining 30  percent of emissions are
from natural sources of methane. Prior to the industrial age anthropogenic methane emissions
were  negligible (Chappellaz et al.  1992).

      While increases in methane's atmospheric concentration in the recent past are largely
attributed to human activities, changes in methane concentrations prior to the industrial age
and spanning back over 150,000 years (as reflected in ice core data) are largely attributed to
changes in emissions from natural sources.3  Emissions  from wetlands, in particular, are
believed to have played a major role in almost doubling atmospheric methane levels twice
   3 Reliable historical data on the atmospheric concentration of methane are available from Antarctic and
Greenland ice cores. The minimum concentration during the last glacial periods (about 20,000 and 150,000 years
ago) was around 350 parts per billion volume (ppbv), and rose rapidly, in phase with the observed temperature
increases to about 650 ppbv during the glacial-interglacial transitions (about 15,000 and 130,000 years ago)
(Lorius et al.  1990; Raynaud et al. 1993; IPCC 1990).  (The current atmospheric concentration of methane is
about 1800 ppbv (IPCC 1990)).
                                                                               Page 1-1

-------
Exhibit 1-1
Sources of Atmospheric Methane
Sources
Natural
Wetlands1
Termites
Oceans
Freshwater
Gas Hydrates
Anthropogenic
Coal Mining, Natural Gas &
Petroleum Industry
Rice Farming
Domesticated Livestock
Livestock Manure
Wastewater Treatment
Landfills
Biomass Burning
Total2
Estimate
(Tg CH4/year)

115
20
10
5
5

100
60
80
25
25
30
40
505
Range

100-200
10-50
5-20
1-25
0-5

70-120
20-150
65-100
10-203
20-25
20-70
20-80
400-610
Source: IPCC 1992.
1 New estimates for wetland emissions are developed in Chapter 2 of this report.
2 Estimation based on observation of atmospheric concentrations rather than sum of individual sources
shown here (Crutzen 1991).
3 Emissions from Livestock Manure reflect, revised estimates. Emissions for all other sources are currently
being updated by EPA (1993)
during this period.4 Increased emissions from gas hydrates have also been proposed as a
reason for the historical rise in atmospheric methane concentration (Nisbet, 1992).  Exhibit
1 -2 shows the atmospheric methane concentrations (bottom line - right scale) and estimated
temperature changes (top line - left scale) during the past 160,000 years as determined on
   4 Chappellaz et al. (1992) estimate that natural wetland emissions rose from 75 Tg CH4/yr during the Last
Glacial Maximum (18,000 years before the present) to 135 Tg CH4/yr during the Pre-lndustrial Holocene (9000 -
 200 yrs BP).
                                                                                 Page 1-2

-------
the ice core from Vostok, Antarctica.  In total, ice core records suggest that a global increase
in temperature of 1 "C causes an increase in atmospheric methane levels of about 50 ppbv
under natural conditions (Raynaud et al. 1993). Assuming an atmospheric lifetime of  10
years, this finding suggests that natural methane emissions increased by about 15 Tg/yr for
each 1 "C increase in temperature.
                                     Exhibit 1-2

      Methane Concentrations and Temperature Changes in the Past 160,000 Years
                        2000
      Ice Core
     Depth (m)
   1500     1000
500
         Methane
          (ppbv)
                                                                AT
                                                                C
                     400
                     300
                       160
120       80      40
  Years Before Present
      (thousands)
  Source: Chappellaz et al. 1990
       The past ice core record, in  addition to other recent research,  suggests there is
potential for the methane emissions from natural sources to increase as climate changes and
further contribute to increasing atmospheric concentrations of methane. The emissions from
several of the natural sources  -- wetlands, gas hydrates, and permafrost - are strongly
governed by environmental variables such  as temperature and precipitation.   As these
environmental variables are altered through climate change, emissions from natural sources
could increase and act as a positive feedback to further fuel increasing global temperatures.
                                                                             Page 1-3

-------
1.3   Current Emissions from Natural Systems

      The largest  source of natural  methane  emissions is  natural wetlands.   Smaller
contributors to natural methane emissions include gas hydrates, permafrost, termites, oceans,
and freshwater. Natural wetlands, gas hydrates, and permafrost regions are likely to play the
largest roles in altering future methane emissions from natural sources as a result of climate
change, and therefore are the focus of this report.

      Natural Wetlands

      Like agricultural wetlands such as flooded rice fields, natural wetlands are significant
sources of methane. They provide a habitat conducive to methane-producing (methanogenic)
bacteria that produce CH4 during  the  decomposition of  organic material.  These bacteria
require environments with no oxygen (a situation promoted by  flooded soils) and  abundant
organic  matter, both of which are characteristics of wetlands (Zehnder 1978).

      The process  by which wetlands emit methane involves the methanogenic bacteria and
methane-consuming (methanotrophic) bacteria. Net emissions to the atmosphere are equal
to methane production by methanogens minus consumption by methanotrophs.

                   Emission  (Net) -= Production (Gross) - Consumption

      More specifically, methane is produced by methanogens in the anaerobic layer of
wetland soil (Exhibit 1 -3).5  Once produced, most of the methane enters the aerobic soil layer
and is oxidized (converted to carbon dioxide and water) by the methanotrophs present in that
layer.6  Methane that is not oxidized  in the aerobic layer eventually reaches the atmosphere
by one of three transport mechanisms. The primary transport mechanisms for the methane
vary among wetland systems with some methane exiting through the plant, some collected
in air bubbles that migrate to the surface, and some undergoing molecular diffusion through
soil and water to reach the atmosphere (Exhibit 1-3).  Notice that in this exhibit the water
table is above the soil surface, which is common in tropical systems, however, in many cases,
particularly in northern ecosystems, the water table may be below the soil surface.

       Currently about 4 percent of the land surface of the earth  is inundated wetlands which
produce and emit methane. These wetlands are concentrated in the high latitudes of the far
north and in the tropics (Exhibit 1-4),  In the north during the thaw season, permafrost below
the soil surface  impedes soil drainage and causes the  flooded conditions  conducive to
methane production. In the tropics high rates of precipitation at any time  during the year lead
to flooded conditions.  Temperate-zone wetlands can exhibit  the characteristics of both
tropical and northern wetlands.
   5 The method by which methane is produced in an anoxic ecosystem is a complex process generally referred
to as an "anaerobic food web". In this process a variety of nonmethanogenic anaerobic microbes attack complex
organics, resulting in the formation of methanogenic substrates. Theses substrates are then metabolized by the
methanogenic bacteria into methane (Cicerone and Oremland 1988).

   6 Studies have shown that methane can be oxidized in the anaerobic layer as well as the aerobic layer
(Cicerone and Oremland 1988).
                                                                               Page 1-4

-------
                                     Exhibit 1-3

                            Methane Emissions Pathways
                                                              CH,
                Anaerobic Soil Layer
                                Plant Transport
                                                CH, Generation
      Wetlands can be forested or unforested.  All forest ecosystems that emit significant
amounts of methane are accounted for this report under natural wetlands. The vast majority
of wetlands are freshwater, as opposed to saltwater. In fact, the only saltwater wetlands that
emit appreciable quantities of methane are saltwater marshes in the temperate zone. These
systems are included in the analysis here for temperate wetlands.

      In addition, moist to dry tundra in the arctic has been identified as a source of methane
emissions. While dry or well-drained tundra has not been considered a wetland in some other
efforts,  it is considered a source in this report due to potential current emissions and due to
the potential for emissions to change as climate  is altered.  The area of well-drained tundra,
is about 5.9 x 1012 m2 (or another 4 percent of the land surface area on earth), for a total
wetland area of  111 x  1012 ma (see Exhibit 1-5).

      Recent studies have estimated that natural wetlands emit about  110 Tg of methane
per year (Matthews and Fung 1987; Aselmann  and Crutzen 1989).   However, substantial
additional  observational data have become available  over the last several years, and new
estimates for emissions for the different regions of natural wetlands are developed from this
data in this report.

Fossil Sources
      The other natural sources of interest, gas hydrates and permafrost, are fossil sources
of methane, meaning that the methane was created in the geologic past, stored in the earth's
                                                                             Page 1-5

-------
                                Exhibit 1-4

                               Wetlands Map
                  "irnzzi

        KEY
        dark shading:   wetland area > 10%
        light shading:   0.5% < wetland area < 10%
        dashed circles:  islands with substantial wetland areas
        (Source: Hofstetter 1983)
 Source: Hofstetter (1983)
Exhibit 1-5 Global Wetland Area
Latitude Region
Tropical Wetlands (30S-20N)
Temperate (45-20N, 30-50S)
Northern Wetlands (above 45 N)
Total Inundated Wetlands
Well-drained Arctic Wetlands
(Tundra)
Total Wetlands
Area (x1011ma)
19
4.1
29.6
52.6
58.7
111
crust, and is only now being released to the atmosphere. Fossil methane can be released by
a variety of activities, including natural gas systems and coal mining operations; however.
                                                                     Page 1-6

-------
these emissions are triggered by human rather than "natural" activity, and thus are classified
as anthropogenic emissions.  Seepage from natural gas reservoirs is a natural fossil source,
but currently there is little information available on the importance of this source.  Thus,
emissions from two sources, gas hydrates and permafrost, represent the larger natural fossil
sources of methane.

      Hydrates.  Hydrates are solids composed of cages of water molecules that contain
      molecules of methane. They are found deep underground in polar regions (continental
      hydrates) and in ocean sediments of the outer continental margin throughout the world
      (oceanic hydrates) (see Exhibit 1-6). A vast amount of methane is stored in hydrates.
      The exact amount is not known but is probably around  107 Tg for oceanic reserves
      and 5x105 Tg for continental reserves. Recent estimates of current emissions from
      hydrates range from 0 to 5 Tg per year (Kvenvolden 1991e; IPCC 1992).

      Historically, major changes in surface conditions of the earth, such as ice ages, have
      caused huge quantities of methane (i.e., thousands of teragrams) to move into and out
      of hydrates. These movements typically take place over the course of thousands to
      tens of thousands of years, and have altered the climate-controlling radiative properties
      of the earth. It has been hypothesized, for example, that the  release of methane from
      hydrates significantly  contributed to the warming which caused the last major ice age
      to come to an end (Nisbet 1990; MacDonald 1990).

      Permafrost. Permafrost methane is created mainly through biological processes and
      trapped in shallow permafrost ice and soil  before it  can  reach the atmosphere.
      Permafrost underlies about a quarter of the land area of the earth. The amount of
      methane stored in permafrost is not well known, but based on the range of observed
      concentrations of methane in permafrost samples, there could be over 5,000 Tg in the
      ice portion of permafrost alone. The average total volume of permafrost is known to
      be decreasing in limited areas. The volume of permafrost that is being lost and the
      volume of methane being released as a  result  have not been calculated.  Large
      quantities of organic matter are also frozen in the permafrost. If thawed, this organic
      matter could increase methane generation  in the active soil layer.  (This possibility is
      accounted for in the chapter on natural wetland emissions.)

      Other Natural Sources

      Several other sources of methane from  non-fossil, natural sources  are known or
inferred to exist. While these sources have not been well researched, it is generally believed
that they are relatively small ones. Furthermore, it has not been hypothesized that emissions
from them are likely to increase as a result of climate change. Therefore, this report does not
discuss potential future emissions from these sources.  Additional  research is necessary to
better understand current emissions from these systems and how they are likely to respond
to changes in climate.

       Termites.  Recent studies  have scaled back  estimates  of methane emissions from
      termites from  10 to 100 Tg/yr to  10 to 50 Tg/yr (IPCC  1992).  Emissions from this
      source are dependent  upon termite population, amounts of organic material consumed
      by termites in various biomass, species differences, and activity of methane-oxidizing
      bacteria (Cicerone and Oremland 1988). While more research needs to be done in this
                                                                             Page 1-7

-------
                                    Exhibit 1-6

                    Locations of Known or Inferred Gas Hydrates
  Source: Kvenvolden (1988)
area, some experts believe that future trends in termite emissions are far more likely to be
shaped by anthropogenic changes in land use (e.g., deforestation  for agriculture) than by
climate change (Nisbet and Ingham, submitted; Bartlett, pers. comm. 1993).

      Oceans and  Freshwaters.   Research conducted  in  the late 60's and early 70's
      established that the surface waters of the world's oceans are slightly supersaturated
      with methane in relation to its partial pressure in the  atmosphere, and therefore are
      currently emitting methane.  In other words, the carrying capacity of the oceans has
      been exceeded. The emission estimate of 5 to 20 Tg/yr in Exhibit 1-1  may be high
      because the atmospheric concentration of methane has probably increased 20% since
      this  range was developed in  1970, thereby  increasing the carrying capacity of the
      oceans (Cicerone and Oremland 1988).  The source of the methane emitted from the
      oceans is not clear.  In coastal regions it could  come from  sediments and drainage
      (e.g., rivers).   It  has also been suggested that methanogenesis occurs within the
      anaerobic gastrointestinal  tracts of marine  zooplankton  and  fish (Cicerone  and
      Oremland  1988).   In  freshwaters,  methane  emissions  can  result  from  the
      decomposition of wetland plants. These emissions are accounted for in this report
                                                                             Page 1-8

-------
      under wetland  emissions.  Additional source(s) of methane from freshwaters are
      believed  to  exist but very little  is known about them.  At this time,  there is  no
      published information that suggests that emissions from oceans and freshwaters will
      increase  in the future.

      Non-Wetland Soil Emissions. There are several ecosystems which do not fall under the
      category of wetlands as used in this report, but are believed to emit methane, at least
      in some areas, over some period of time. An example might be relatively well-drained,
      boreal forests that are water-saturated after snow-melt (Bartlett, pers. comm. 1993).
      However, emissions from these systems are not considered significant because they
      have not  been included in any of the published studies of methane sources (Matthews
      and Fung 1987; Cicerone and Oremland 1988;  Fung et al. 1991).
1.4   Future Emissions from Natural Sources

      Emissions from natural sources are  strongly influenced by environmental variables,
such as soil temperature and inundation (water level),  which will be affected by climate
change.  Natural methane emissions may, therefore, represent a positive climate feedback
process.  Climate change induced  by humans could trigger the release of more greenhouse
gases from natural  systems.  The implications of this effect are twofold:

            The rates and magnitude of future climate change would increase.

            The problem of controlling emissions of the greenhouse gases, so as to reduce
             the adverse effects of climate change, would  be greatly exacerbated.

      Feedback processes are natural systems that have the ability to amplify or dampen the
initial change in radiative forcing caused by increasing concentrations of greenhouse gases.
There are two types of feedbacks:  (1) geophysical feedbacks are part of the internal physical
dynamics  of climate; and (2) biogeochemical feedbacks deal with the earth's biology and
chemistry.

      The more important geophysical climate feedbacks -- water vapor, clouds, and sea ice
albedo - are  accounted for in current efforts to predict  climate change through the use of
Global Circulation Models. For example, the water vapor feedback is modeled by estimating
the additional water vapor that the atmosphere can hold as  it warms.  The additional water
vapor, which is a greenhouse gas,  amplifies the initial warming. This positive feedback acts
to approximately double the initial warming  of the atmosphere that would result from human
activities alone (EPA 1990). Biogeochemical feedbacks, such as changes in natural methane
emissions, ocean CO2 uptake, and  vegetation albedo, have generally not been included in the
climate models  (Lashof 1989).

      The ability to quantify the impact of biogeochemical feedbacks is limited by the current
understanding  of  these  systems.7  However, these  feedbacks could  contribute  to a
substantially larger potential warming than  is now predicted by climate models. As Exhibit
   7 Analyses of ice core data indicate that natural sources of methane may have provided significant feedback
to climate changes in the past (Raynaud et al. 1993).
                                                                             Page 1-9

-------
                                   Exhibit 1-7

      Climate Feedbacks to Doubled CO2 (Based on 1.5 to 5.5 Degree Sensitivity)
                 7 -
       Equilibrium
      Temperature    _
      Change (C)

                 3 -
                 2 -
                  1 -
                                                 5.5

                                                           '////
                                                           / / / / .'
                         1.25
                                                           w,
                                                 1.5
                          T
                        No
                     Feedbacks
   I
  Water
  Vapor
Feedback
    I
   Geo-
 I
All
 physical    Feedbacks2
Feedbacks1
  1 Includes: Water Vapor, Ice and Snow, Clouds
  2 Includes: Geophysical  Feedbacks and Biogeochemical Feedbacks (Natural Methane,
  Ocean Effects and Vegetation Effects)
  Sources: Lashof (1989); EPA 1990
1-7 illustrates, climate models generally predict that temperatures will rise approximately
1.2C (within 50 to 100 years) as a result of the emissions from human activities alone, with
a doubling of greenhouse gases in the atmosphere (CO2 equivalents). When the water vapor
feedback is added, temperatures are predicted to rise to 1.7 to 2.6C. When ice, snow, and
clouds are added, the models predict temperatures to rise by 1.5 to 5.5C (EPA 1990). The
temperature rise  if all feedbacks  --  geophysical and biogeochemical  are  included is highly
uncertain, but it is likely to be greater than the temperature  rise when  only geophysical
feedbacks are considered (Lashof 1989).
1.5   Overview of this Report

      In the Clean Air Act Amendments of 1990,  Congress requested that the U.S.
Environmental Protection Agency investigate the extent to which natural methane emissions
                                                                          Page 1-10

-------
may provide a positive feedback to climate change, and to prepare a report to Congress for
submission no later than two years after enactment of the Act on:

      methane emissions from  biogenic sources such  as (a)  tropical,  temperate, and
       subarctic forests, (b) tundra, and (c) freshwater  and saltwater wetlands; and

      the changes in methane emissions from biogenic sources that may occur as a result
       of potential increases in temperatures and atmospheric concentrations  of  carbon
       dioxide.

       This report was prepared to fulfill the above request.8 The report reviews the state-
of-the-art for estimating emissions of methane from natural systems to the atmosphere and
understanding the factors controlling emissions from these systems. The report also attempts
to identify a set of  possible  changes in emissions from  natural systems in  response to a
changing climate.

       This report is divided into two major sections:

       Chapter 2: Natural Wetlands. This chapter reviews methane emission data collected
       from wetlands worldwide, provides average  emission  rates for different wetland
       ecosystems,  develops global  emission estimates,  presents some possible  future
       scenarios,  and discusses  further research that  could  reduce  uncertainty in the
       developed estimates.9

       Chapter 3: Fossil Sources. This chapter - divided into gas hydrates and permafrost -
       -  describes the different systems, provides available estimates of current emissions,
       presents scenarios for future emissions, and discusses the major uncertainties and how
       they could be resolved.
   8 The congressional request also asked for interagency coordination of the development of the methane
reports.  This coordination was accomplished through formation of an Interagency Working Group on Methane
which met quarterly to discuss important issues.

   9 Much of the information for this chapter was developed through a grant from the EPA to the University of
New Hampshire and the efforts of R. Harriss, K. Bartlett, and P. Grill. (A separate article is being submitted by
these researchers for publication in a scientific journal.) In order to assess the potential impacts of climate change
on natural wetlands, a workshop was held at the University of New Hampshire on March 10 and 11,1992, as part
of the EPA grant. A list of the wetlands researchers who participated in this workshop is provided in Appendix
A.
                                                                                Page 1-11

-------
                                     CHAPTER 2

                                NATURAL WETLANDS
       Natural wetlands is a general term  for a broad  range of unique ecosystems.  To
determine the global annual emissions of methane from wetlands a classification system is
developed by which wetlands ecosystems with similar methane emissions characteristics can
be reasonably aggregated into general wetland types.  The areas, average emission rates, and
annual period of emissions for these wetland types are then calculated.

2.1    Background - Wetlands Classification

       Wetlands  are commonly divided by  latitude into three major  regions:  tropical,
temperate, and northern. Fundamental differences exist in the physical and climate processes
that characterize tropical and northern  systems;  temperate  systems  can  exhibit the
characteristics of either tropical or northern systems.

       In the tropics, high rates of precipitation lead to flooded conditions in low-lying areas,
such as river basins. These high precipitation rates, coupled with the high temperatures and
solar radiation rates in  the tropics, are conducive to plant growth and decay. In  areas  of
inundation, this decay process produces methane, which can be released into the atmosphere.
Thus,  methane is emitted year-round in permanently inundated  wetlands and during the
inundation period in  wetlands inundated  only seasonally.

       In the northern region, wetlands are common because permafrost exists just below the
soil surface and inhibits the drainage of  moisture from the upper soil layer.  While recently
discovered evidence suggests that some  methane may be emitted from northern wetlands in
the winter, northern wetlands generally only emit methane during the summer, the thaw
season. These wetlands also differ from tropical ones  in that they generally contain large peat
deposits (partially decomposed plant residue).10

       The division of wetlands into tropical, temperate, and northern is a generalization that
alone does not adequately characterize wetlands in terms of methane emissions, since fluxes
vary widely within these regions as determined  by vegetation or habitat type. Therefore,
additional subdivisions  by vegetation type have been developed.  These subdivisions have
been chosen to maintain relatively homogeneous systems and assist in developing  methane
emission estimates.  These further subdivisions are discussed below in terms of their major
characteristics, their relative emissions of methane (discussed in detail in section 2.2), and
their areal extent.
   10 The accumulation of peat indicates that at a particular location and moment in time the rate of plant matter
deposition is greater than the rate of anaerobic decomposition (e.g., that some variable other than organic matter
is the limiting variable in the decomposition equation).  The presence or absence of peat does not necessarily
indicate the size of methane emissions.  However, it is important information for  developing future emission
scenarios because peat is an existing source of carbon which can be converted to methane if certain conditions
are met in the future.
                                                                                Page 2-1

-------
2.1.1  Tropical Wetlands

       Wetlands in  the tropics (roughly 20 N to  30 S) are generally characterized by
relatively high methane emission fluxes. The high methane emission rates result from a rapid
cycle of plant growth and decomposition due to high temperatures and elevated levels of solar
radiation.   However, emissions vary substantially for different ecosystems.  For example,
ecosystems with limited vegetation and in alluvial (flowing water) settings have less available
carbon for decomposition and relatively low emissions.  To define homogeneous systems for
estimating methane emissions, tropical wetlands can be divided into three distinct categories
each with a specific area of coverage:11'12

       Non-forested Swamps.  Non-forested  swamp wetlands generally have the highest
       emission rates of the tropical  systems.  These wetlands  are found  mostly  in  the
       Amazon floodplain  and contain several species  of  grass which have  adapted to
       seasonal inundation. These species have high growth rates over much of their life cycle
       in  order to maintain access to sunlight.  Grasses eventually  break away from  the
       submerged soils and float freely in large mats. Non-forested swamps are also generally
       not alluvial and are  peat-poor.  Non-forested swamps  are estimated to extend over
       7.9x1011 m2, comprising  40% of tropical wetlands, and 15% of global inundated
       wetlands.13

       Flooded Forests. Flooded forests, sometimes referred to as forested swamps, tend to
       emit more methane than  unvegetated open water,  but less  methane than grassy
       wetlands. They can be seasonally or permanently inundated.  For example, extensive
       seasonal wetland areas may occur in the  flood plains of large rivers such as  the
       Amazon.  Flooded forests are usually characterized by  minimal peat deposits and by
       standing water, but  they are occasionally rich in peat and/or alluvial. Flooded forests
       are estimated  to cover 10.4x1011  m2 in the tropics, comprising 55%  of tropical
       wetlands and 20%  of global inundated wetlands.

       Open Water. Open water wetlands are the smallest category of tropical wetlands  and
       have the lowest emission rates. They generally lack vegetation and can be seasonally
   11 Areas for the different ecosystems were derived from Matthews and Fung (1987), who divided wetlands
into 10 latitude bands, and areas within each band, into 5 wetlands classifications: forested bog, non-forested
bog, forested swamp, non-forested swamp, and alluvial wetland formations.  Bogs are usually rich in peat, while
swamps are peat-poor. Unlike bogs and swamps, alluvial formations are characterized by flowing surface water.
The total wetland area from Matthews  and Fung (1987) is preserved in this report and simply categorized
differently according to this classification system. The following correlations have been made between tropical
wetland types used by Matthews and Fung (1987) and the breakdown described above:
 Non-forested swamps (in this report) = non-forested swamps  + non-forested  bogs + one third of alluvial
formations (in Matthews and Fung).
 Flooded Forests = forested swamps + forested bogs + one third of alluvial formations.
 Open Water  = one third of alluvial formations.

   12 This classification system has been developed to correspond to the current set of measurements from
tropical regions. As new measurements are taken in different systems, this classification system may be modified.

   13 The area of each wetland category is compared to global inundated wetlands rather than total wetlands
(inundated + well-drained arctic tundra) because well-drained tundra has not been considered a wetland in several
previous studies, and therefore this comparison is more  useful for cross-referencing.
                                                                                  Page 2-2

-------
      or permanently inundated. Open water wetlands are estimated to cover 0.5x1011 m2,
      which is 3% of tropical wetlands and 1 % of global inundated wetlands.

2.1.2  Temperate Wetlands

      While a wide variety of wetland types are found in temperate to subtropical regions
(45-20 N and 30-50 S), these regions account for only about 7% of the total inundated
wetland area (4.1x1011m2). Temperate wetlands generally produce much less methane per
unit area than tropical systems, in part due to lower temperatures and lower solar radiation.
For this analysis, temperate wetlands are grouped into the four categories used  by Matthews
and Fung (1987) to describe this region:

      Forested Bogs.   Forested Bogs generally have the highest emission  rates  of the
      temperate systems. They are dominated by shrub wetlands (evergreen and drought-
      deciduous) and high-latitude, temperate, boreal forest/woodland/shrub  wetlands.  In
      general, bogs have significant peat  deposits.  The areal extent of these bogs  is
      estimated at 1.0x1011 m2, or 2%  of global inundated wetlands.

      Forested Swamps.  Forested swamps in the temperate zone are  wooded wetlands,
      usually with minimal peat accumulation. Average emissions are considerably less than
      those of forested bogs.14  The area is estimated to be 1.3x1011 m2, or 2%  of global
      inundated wetlands.

      Non-forested Swamps.  Non-forested swamps are  peat-poor, inundated grasslands
      covering an area of 1.3x1011 m2, or 2% of global inundated wetlands.  They tend  to
      emit about as much methane as forested swamps.

      Alluvial Formations.  Alluvial formations are the smallest methane producers  of the
      temperate  systems.  In  the temperate zone, alluvial formations are  usually cold-
      deciduous, alluvial forests. The area is estimated to be 0.4x1011 m2, or less than 1 %
      of global inundated wetlands.

2.1.3  Northern Wetlands

      A number of factors distinguish different ecosystems in the northern wetlands.  These
factors include soil moisture and the presence of winter emissions.  To address these factors,
northern wetlands are  generally  divided  into arctic  (above 60N) and  boreal (45-60N)
wetlands.  Since soil moisture is a major factor controlling methane emissions, arctic and
boreal wetlands are further  subdivided  into "flooded,"  "well-drained," and "mixed" (both
flooded and well-drained soils) moisture classes.  Well-drained wetlands are areas where the
water table is below the surface, and  there is no standing water at the soil surface. Finally,
at several  sites both in the arctic and  boreal regions, small lakes and ponds appear  to
contribute significantly to emissions.
   14 This difference in emission rates is based on a limited sample size and may not be statistically significant;
see section 2.2.2
                                                                             Page 2-3

-------
      Given these factors, northern wetlands may be divided into eight categories:
                     ARCTIC WETLANDS    BOREAL WETLANDS
flooded
well-drained
mixed
lakes
flooded
well-drained
mixed
lakes
      The major emission differences between these ecosystems are (1) flooded wetlands
have significantly higher emissions than well-drained ones, with mixed wetlands falling
somewhere in between; (2) lakes appear to be significant sources of methane, but the data
set is very limited; and (3) winter emissions are believed to be higher in the boreal region,
although there is little data. Arctic and boreal latitudes appear to have similar emission rates
at other times.

      Data on areal extent is not specifically available for all eight of these wetland types.
Therefore, the  following  three categories (shown in bold on the table above)  are used to
represent all northern wetlands: (1) flooded arctic wetlands, (2) flooded boreal wetlands, and
(3) well-drained arctic wetlands. These three  systems are believed  to represent the vast
majority  of northern wetland areas, and therefore, the global  emissions results should not
suffer significantly from the use of  these three  systems to represent all northern wetlands.
However, more accurate estimates of northern wetland emissions can be made in the future
if the specific area of all eight types is defined.

      Flooded Arctic Wetlands. In the arctic, wetlands vegetation  is limited primarily to
      grasses, sedges, and low shrubs. Since these wetlands are generally not associated
      with rivers, small variations in miicrotopography commonly create differences in inunda-
      tion and vegetation.15   Inundated wetlands in the arctic are estimated  to  cover
       14.7x1011 m2, which  is 50% of northern wetlands and 28% of global inundated
      wetlands (Matthews and Fung 1987).16

      Flooded Boreal Wetlands.  Moving south from the arctic into the boreal region, trees
      become more common and wetlands become more diverse.  Common wetlands in the
      boreal zones include  bogs and  fens.  Bogs are peat-producing  wetlands that are
      characterized by their isolation from surface-water sources. They receive water and
      nutrients only from rainfall, a condition termed ombrotrophy. These wetlands are quite
      acidic and poor in  nutrients, and vegetated by plants adapted to these rigorous condi-
      tions.    Fens,  on  the  other  hand, are commonly in  contact with  surface waters
   16 On a small scale, the floodplain of a river is generally a homogeneous system in terms of topography,
inundation and vegetation.  However, non-riverine systems - which are not smoothed by the action of a river -
can exhibit uneven topography leading to flooded and non-flooded areas in close proximity.

   16 In deriving this area, Matthews and Fung (1987) included areas that are referred to above as mixed arctic
and arctic lakes.
                                                                               Page 2-4

-------
      (minerotrophy)  and are more alkaline,  have higher nutrient levels, and have  more
      diverse vegetation.  The area of flooded boreal wetlands is estimated at 15x1011 m2,
      comprising 50% of northern wetlands  and 28%  of global  inundated  wetlands
      (Matthews and Fung 1987).17

      Well-drained Arctic.  These wetlands (also called dry tundra) are characterized by
      treeless, vegetated plains and a water table that is generally below but near the soil
      surface.  Dry tundra area covers an estimated 58.7 x1011 m2, which is larger than the
      entire area of global inundated wetlands.  (This number is derived by subtracting the
      area of high latitude inundated wetlands (Matthews and Fung 1987) from the total
      area of tundra (Matthews 1983).)
2.2   Review of Emission Measurements

      This section provides a state-of-the-art review of methane emission measurements.
Using the classifications  developed above, this section also provides an estimate of the
average methane emission rate from each of the major wetland types.

      During the past decade, the number of studies on methane production and release from
wetland environments has increased substantially, especially in  North America, making it
possible to develop more reliable average emission rates for the major wetland types.  Since
methane emissions can vary by orders of magnitude in relatively small time- and space-scales,
extrapolating emissions from  flux  measurements can be an  uncertain process.  This
characteristic variability in fluxes also increases the difficulty in making cross-study and cross-
ecosystem comparisons as these depend  on  the degree to which measurement sites are
representative of larger systems.  It is therefore encouraging to note that there is reasonable
agreement in the  results  of  the various  measurement programs to date. Moreover,  these
measurement programs include a variety of measurement techniques spanning spatial scales
from less than 1 m2 to hundreds of km2 (Fan  et al. 1992; Bartlett et al. 1992; Ritter et al.
1992; Roulet, pers. comm.).

2.2.1 Tropical Measurements

      Emission measurements from the tropics are generally high in comparison with those
from the temperate and  northern  regions.  They are also variable,  which is  probably a
consequence of CH4 release  by bubbling, as reported in various studies (Bartlett et al. 1988;
Crill et al. 1988; Bartlett et al. 1990; Devol  et al. 1990; Keller 1990). Ebullition, or bubbling,
is one of three pathways  by which methane can escape from wetlands to the atmosphere.
Since bubble release is sporadic and can emit large amounts of gas, measurements taken with
and without bubbling are  dramatically different.

      Tropical emission measurements have been made at several sites in the Amazon, and
sites in the Orinoco river flood plain in South America, Panama, and the Congo river basin in
Africa (see Exhibit 2-1). The CH4 flux data from these sources is summarized in Appendix B.
All of the data included in Appendix B have been published in the last four years.
   17 In deriving this area, Matthews and Fung (1987) included areas that are referred to in this report as mixed
boreal and boreal lakes.
                                                                             Page 2-5

-------
Exhibit 2-1
Measurements in Tropical Regions
Summarized in this Report
Tropical Site
Amazon
Orinoco
Panama
Congo
References
Devol et al. 1988
Bartlett et al. 1988
Bartlett et al. 1990
Devol et al. 1990
Wassmann et al. 1992
Smith and Lewis 1992
Keller 1990
Tathy et al. 1
992
      The tropical emission measurement data from Appendix B can be aggregated into the
three wetland types defined for the tropics (open water, non-forested swamps, and flooded
forests; see Exhibit 2-2) with the following results:
      Non-forested Swamps. Emissions from non-forested swamps are generally higher than
      the emissions from other tropical systems and cover a wide  range. In the Amazon,
      average fluxes from non-forested swamp systems range from  131 to 390 mg CH4/m2"
      Id, with most values at about 200 mg CH4/m2/d. Data from the Orinoco River appear
      to be an order of magnitude lower (30 mg CH4/m2/d), suggesting that there may be
      significant differences between the two systems.

      Flooded Forests.  Emissions from  flooded forests  tend to be somewhat lower than
      those from non-forested swamps but also cover a wide range, from 7 to 230 mg
      CH4/m2/d in the Amazon. Values  from inundated African forests may be somewhat
      lower than  those from the Orinoco (106 and 174-307 mg CH4/m2/d, respectively),
      while those from a site in Panama appear to be somewhat greater (346 mg CH4/m2/d).

      Open water. Open-water sites generally have lower emissions than sites with vegeta-
      tion (floating  macrophytes (mats) and  flooded forests).  In the Amazon, average
      methane fluxes from open water span a relatively narrow range, from 27 to 88 mg
      CH4/m2/d.  Open water flux measurements from Panama cover a broader range, from
      54 to 967 mg CH4/m2/d.  However, since most of the Amazonian open-water sites
      studied were  in water depths similar to the deeper water Panamanian sites (greater
                                                                          Page 2-6

-------
      than 5 meters) with lower emissions, these data sets may actually be consistent.
      High-water data from seven lakes on the Orinoco River flood plain are somewhat lower
      (an average of 7.5 mg CH4/m2/d) but are consistent with the Amazonian data.

      In  order to  calculate average  emission  rates  from the  database of  tropical
measurements, a weighted average of high-water and low-water measurements is calculated
in  order to account for annual variability.  This high-low average emission rate is then
averaged with those measurements from sites observed year-round (see Exhibit 2-2).
Exhibit 2-2
Average Emission Rates from Tropical Wetlands
(Derived from Appendix B)
HABITAT
Non-forested Swamps
Flooded Forest
Open Water
AVERAGE EMISSION RATE
(mg CH4/m2/d)
233
165
148
2.2.2 Temperate Measurements

       All of the data collected in this climatic region are from wetland sites in the United
States.  In general, the data fall into two broad vegetation types: forested swamps and non-
forested swamps (both saline  and freshwater). These vegetation types correspond to the
dominant types of U.S.  wetland areas.  Appendix C presents CH4 emission data from
wetlands in the temperate to subtropic zones, roughly from 45 to 25 N.

      The observed emission rates range widely, from isolated negative fluxes (consumption
of atmospheric CH4) to emissions more than three orders of magnitude higher: -7.9 to 3,563
mg CH4/m2/d.  Emission rates for the different ecosystems are:

      Forested Bogs.  Emission rates  for forested bogs appear to be the highest  for
      temperate systems, although the number of emission measurements is limited. The
      available measurements were averaged to derive a mean temperate bog flux of 135
      mg CH4/m2/d (number of sites, n, = 5) (Yavitt et al. 1990; Crill, unpublished data).

      Forested and Non-forested Swamps.  These  systems appear to have  emission rates
      about half as large as  forested bogs.  An average emission rate for temperate forested
      swamps was calculated to be 75 mg CH4/m2/d (n = 18), and for non-forested swamps,
      70 mg CH4/m2/d (n = 25) (see Appendix C).
                                                                           Page 2-7

-------
      Alluvial Formations.  Emissions from alluvial formations were estimated from brackish
      and fresh open water (DeLaune et al. 1983; Appendix C); lakes in the Okefenokee
      Swamp (Bartlett, unpublished data);  and flood plain cypress swamps (Harriss and
      Sebacher 1981). The emissions average 48 mg CH4/m2/d.
2.2.3 Northern Measurements

      Measurements from the northern wetlands span more than three orders of magnitude,
from less than 1 to roughly 1,940 mg CH4/m2/d, and include a wide variety of vegetation,
moisture, and  soil  types.  A single  flux of 12,068 mg  CH4/m2/d, roughly  an  order of
magnitude greater than all others at this site, is reported from an Alberta beaver pond.
Appendices D, E, and F list reported flux measurements from northern wetlands, covering a
latitudinal range of 45 to 70N. These measurements may be divided between arctic and
boreal measurements as described below.

      Arctic Measurements

      Measurements in subarctic and arctic tundra (above 60 N)  are mainly from three
regions in Alaska:

            Fairbanks,  in the interior of the state, where  the only  annual, multi-year
             measurements have been made (Whalen  and Reeburgh 1988;  Whalen and
             Reeburgh, in press).

            The coastal plain on the North Slope, where at least four separate investigators
             have  made measurements encompassing  both large and small spatial scales
             (Sebacher et al.  1986; King et al.  1989; Whalen and Reeburgh 1990a;
             Morrissey and Livingston 1992).

            Coastal  tundra on the delta of the Yukon and Kuskokwim  Rivers, where
             investigators have made measurements on a variety of scales using an array of
             measurement techniques, including tower and aircraft measurement, during the
             NASA Arctic Boundary Layer Expedition (ABLE 3A) (Bartlett et al. 1992; Fan
             et al. 1992; Ritter et al. 1992).

      Within each of these three regions, the agreement in methane flux data was fairly good
among measurements in similar habitats (e.g., North Slope region: Sebacher's wet coastal
tundra had average emissions of 119 mg CH4/m2/d; Whalen & Reeburgh's wet tundra had
average emissions of 90 mg CH4/m2/d; Livingston & Morrissey's meadow tundra had average
emissions of 64.4 mg CH4/m2/d; wet tundra  = 100 mg CH4/m2/d; very wet tundra = 254
mg CH4/m2/d).  Agreement also appears fairly good among these three regions.

      The only arctic flux measurements not from Alaska were  made at a series of sites in
a  Swedish  mire, ranging from drier  raised  areas  (ombrotrophic) to wetter depressions
(minerotrophic) (Appendix  D).  The  average fluxes from hummocks and higher areas in
Sweden are quite similar to those from comparable Alaskan sites. The average flux from
wetter sites in Sweden (360 mg CH4/m2/d) was higher than  emissions from  wet Alaskan
sites, but this result was largely driven by high measurements on a single date.
                                                                           Page 2-8

-------
      Boreal Measurements

      In the boreal zone (roughly 45-60 N), measurement sites fall primarily into two areas:

           Northern  Minnesota (in and around the Marcell Forest  and the Red  Lake
             Peatland),

           Southeastern/southern Canada  (Schefferville area and the Hudson Bay Low-
             lands).  Small- and large-scale measurement techniques were used in this area.
             For example,  in 1990, an integrated flux measurement campaign involving
             chamber enclosure (small-scale) measurements, and eddy correlation flux (large-
             scale) measurements taken from a micrometeorological tower and from several
             aircraft was  conducted  in the Schefferville  Canada  area (NASA's ABLE
             SB/Canadian NOWES).

Emissions from more western sites in Alberta, Canada have also been recently examined (Vitt
et al. 1990).

      A comparison  of measurements within one boreal region, made at different times by
a variety of investigators suggests that although there may be significant spatial and temporal
variability, average measurements generally agree, in part because flux measurements have
relatively high variance.  In  Marcell  Forest, for example, standard errors range from 3% to
35% of  means (Harriss et al. 1985;  Crill  et al. 1988; Dise, submitted (a)).  However, signifi-
cant differences appear to exist between the two major regions, northern Minnesota  and
southeastern/southern Canada. It is currently unclear why these regional differences occur,
but they suggest the difficulty in extrapolating flux measurements to other regions, even those
with similar vegetation and climate regimes.

      In summary, average emissions from flooded arctic soils (96 mg CH4/m2/d) appear to
be approximately equal to those from flooded boreal regions (87 mg CH4/m2/d). However,
differences within the boreal sites measured complicate this conclusion, and emissions during
winter months in the  boreal zone may be higher. Well-drained tundra soils  in the arctic most
frequently appear to be small sources, 0.6-11 mg CH4/m2/d, with sporadically occurring nega-
tive fluxes (consumption of atmospheric methane) generally between -0.5 and -3 mg CH4/-
m2/d (Whalen and Reeburgh 1990a; 1990b; Fan et al.  1992; Bartlett et al. 1992; King et al.
1989).  The  average  flux from well-drained arctic tundra is 7 mg CH4/m2/d.

      Emission data available for northern wetlands are summarized in Exhibit 2-3. Sufficient
data were available to derive average emission rates from almost all eight  of the unique
northern systems.  However, it is currently possible to measure the areal extent of only three
of these systems: flooded arctic, flooded boreal, and well-drained arctic.  Therefore, global
emissions are calculated using the average emission rates from these three systems (shown
in bold face). The area used to represent flooded boreal wetlands is assumed to include the
areas of mixed boreal and boreal lakes.  Similarly, flooded  arctic includes mixed arctic and
arctic lakes.  Thus, the only area that may not be accounted for is well-drained boreal, which
is probably insignificant in terms of  global emissions (Patrick Crill, pers.  comm.  1992).
                                                                            Page 2-9

-------
Exhibit 2-3
Average Emission Rates from Northern Wetlands
(Derived from Appendix D; Flux units: mg CH4/m2/day)
Moisture
Class
Flooded
Well-
drained
Mixed1
Lakes
Region II Avg.
|| Flux
Arctic
Boreal
Arctic
Boreal
Arctic
Boreal
Arctic
Boreal
96
87
7
10
21
-
51
172
Std. Error
of Mean
21
18
2.0
2.7
9.6
-
16
75
Number of
Studies
28
49
14
5
3
0
6
8
Range
2.8 - 360
0-664
0.6 - 29
3.3-21
1.6-31
-
3.8- 112
12-518
Includes a mixture of flooded and well-drained soils within the
vegetation grouping.
2.3   Global Emissions

      Global annual methane emissions from natural wetlands is the sum of the emissions
from the individual ecosystems. To calculate these emissions, it is necessary to know the
area of the ecosystem, its average daily emission rate, and the active period of emissions over
the course of a year.  With this  information, the following equation can be used to estimate
the contribution to global emissions from a particular wetlands system (Matthews and Fung
1987; Aselmann and Crutzen 1989):

              Annual Emissions = Area x Emission Rate x Emission Period

where annual emissions is Tg CH4/yr, area is m2, emission rate is mg CH4/m2/day, and period
is days/year.

2.3.1 Global Tropical Emissions

      Globally, tropical wetlands are estimated to emit about 66 Tg  of methane annually.
Non-forested swamp wetlands contribute 34 Tg CH4/yr;  flooded forest wetlands contribute
31 Tg/yr; and open water wetlands contribute 2 Tg/yr. These annual estimates are based on
the information summarized in Exhibit 2-4. The average emission period in the tropical region
is estimated to be 180 days/year because this is roughly the length of the annual  wet season
(Matthews and Fung 1987).
                                                                           Page 2-10

-------
Exhibit 2-4
Global Tropical Methane Emissions
Ecosystem
Non-For. Swamps
Flooded Forests
Open Water
TOTAL
Emission Rate
(mg CH4/m2/d)
233
165
148

Area
(1011m2)
7.9
10.4
0.5
19
Emission
Period
(days)
180
180
180
180
Annual
Emissions
(Tg/yr)
34
31
1.7
66
      The methane emission estimate of 66 Tg/year from tropical wetlands is significantly
higher than estimates derived by other researchers.  This difference is due to the use of the
current database of measurements, which indicates that the average emission rates from this
region are considerably higher than previously believed. Using a substantially smaller data
base than was available for the present analysis, Matthews and Fung (1987) calculated that
tropical wetlands contribute 28 Tg CH4/year. Using different data sources for wetland areas,
emission season assumptions, and a data base on fluxes that was larger than the one used
by Matthews and Fung but smaller than the current data base, Aselmann and Crutzen (1989)
estimated that tropical wetlands release about 42 Tg/yr.

2.3.2 Global Temperate Emissions

      Temperate wetlands are estimated to release 5  Tg  CH4/yr.  This annual estimate is
based on the emission rates, area and emission periods listed in Exhibit 2-5.  Annual changes
in both temperature and inundation  determine the emission period in the temperate zone,
where seasonal as well as permanent wetlands are present. On average, the emission period
is estimated to last for 150 days each year, with the exception of those wetlands  between
30 N and 20 N, where it is estimated to last for 180 days. This period corresponds roughly
to the annual period of warm temperatures.

2.3.3 Global Northern Emissions

      Total emissions for flooded  soils in the  arctic and boreal regions are calculated to be
14.1 Tg/yr and 19.5 Tg/yr, respectively. A global flux of 4 Tg/yr is calculated from dry (well-
drained) tundra. The total northern wetlands contribution to global emissions is therefore 38
Tg/yr.  These annual estimates are based on the information listed in Exhibit 2-6.

      In these extreme environments, temperature is the primary factor creating seasonality.
The assumed emission periods are 100 days for areas above 60 N latitude and 150 days for
areas between 45 N  and 60 N.  These periods correspond roughly to the  annual thaw
season, as winter fluxes are assumed to be zero.
                                                                           Page 2-11

-------
Exhibit 2-5
Global Temperate Methane Emissions
Ecosystem
Forested Bogs
Forested Swamps
Non-forested Swamps
Alluvial Formations
TOTAL
Emission Rate
(mg CH4/m2/d)
135
75
70
48

Area
dO112)
1.0
1.3
1.3
0.4
4.1
Emission
Period1
(days)
150, 180
150, 180
150, 180
150, 180
150, 180
Annual
Emissions
(Tg/yr)
2.1
1.6
1.5
0.3
5.4
1 Temperate emissions are calculated using a period of 1 50 days, except for temperate
wetlands between 20 N and 30 N, for which a period of 180 days is used.
Exhibit 2-6
Global Northern Methane Emissions
Ecosystem
Boreal (Flooded)
Arctic (Flooded)
Arctic Tundra
(Well-drained)
TOTAL
Emission Rate
(mg CH4/m2/d)
87
96
7

Area
(1011m2)
15.0
14.7
58.7
88.4
Emission
Period
(days)
150
100
100

Annual
Emissions
(Tg/yr)
20
14
4
38
      This estimate for the northern contribution to global emissions is somewhat smaller
than earlier estimates for this region.  In 1987,  based largely on northern emissions from
Sebacher et al. (1986), Matthews and Fung (1987) estimated that northern wetlands (50-70
N) made a major contribution to atmospheric methane, calculating an annual flux of 62 Tg,
or roughly  60%  of the total emissions from all wetlands. A recent recalculation of the global
contribution from wetland ecosystems (Bartlett et al. 1990) based on the Matthews and Fung
wetland areas and model but using a larger flux data base, suggested that global emissions
from northern high-latitude areas may be lower than those estimated by Matthews and Fung,
but still roughly  39 Tg/yr.  Aselmann and Crutzen (1989) also found that northern emissions
                                                                           Page 2-12

-------
are somewhat lower than first thought and estimated that the region releases about 24 Tg/yr.
Based on their data from Fairbanks, Whalen and Reeburgh calculated global emissions from
areas north of 50 N ranging from 28 to 102 Tg/yr.  If uncertainty in vegetation cover types
is included, the range in emissions encompasses nearly one order of magnitude 13.7 to 134.5
Tg/yr (Whalen and Reeburgh 1992).

2.3.4 Summary of Global Emissions from Wetlands

      The contribution of methane emissions from all wetlands is 109 Tg (see Exhibit 2-7).
Of this total emissions, 60% comes from tropical wetlands.  While tropical wetlands account
for a considerably smaller area than northern wetlands, their emission rates are substantially
higher and they are active for a longer period each year. Temperate wetlands have emission
rates of approximately  the same  magnitude as  inundated northern systems,  and longer
emission periods, but the areal extent is small enough that they account for only 5% of global
wetland emissions.  Northern wetlands (including dry tundra) account for the  majority of
global wetland area, but  given the very low emission rate of dry  tundra, northern wetlands
account for only 35% of total  wetland  emissions.

      While the regional totals calculated here are somewhat different from other studies,
the overall total is similar to other global emission estimates. These totals have ranged from
80 to 115 Tg/yr in recent studies (see  Exhibit 2-8).
2.4    Effects of Environmental Variables

       To investigate future methane emissions from wetlands, it is necessary to understand
the factors that control methane production and release from different wetland ecosystems.
This section discusses what is currently known about the major environmental variables that
influence methane emissions in these different systems.  This section also presents the results
of a 1992 workshop held to discuss these environmental variables and the rankings that were
developed in the workshop to characterize the importance of the variables in the different
systems.18

       Methane  fluxes from  natural  wetlands  have  been linked to a  wide  array  of
environmental variables.  However, not all  of these variables  are expected to be  altered
significantly  by climate  change.   Therefore,  the  following  discussion   will focus  on
environmental variables that could play a  major role  in  influencing future emissions.
Furthermore, because  the temperate zone makes a small contribution to global wetland
emissions compared to other regions, discussion is concentrated on the tropical and northern
zones.

       To investigate the effect of changing climate, it is necessary to examine the potential
effect of the important environmental variables on the emission  rate itself, the wetland area
and  the emission period.  Separating variables  into those affecting the emission  rate as
   18 A workshop on variables controlling future methane emissions from wetlands was held at the University
of New Hampshire on March 10 and 11 1992 through a grant by EPA's Office of Air and Radiation. The workshop
brought together experts in fields related to trace gas emissions from natural systems. Workshop participants are
listed in Appendix A.
                                                                             Page 2-13

-------
Exhibit 2-7
Global Annual Methane Emissions from Natural Wetlands
Ecosystem
Tropical Wetlands
Non-Forest Swamps
Flooded Forests
Open Water
Temperate
Forested Swamps
Non-Forest Swamps
Alluvial Formation
Forested Bogs
Northern Wetlands
Boreal
Arctic
Well-drained Tundra
Total Wetlands
Emission Rate
(mg CH4/m2/d)

233
165
148

75
70
48
135

87
96
7

Area
(x1011m2)
19
7.9
10.4
0.5
4.1
1.3
1.3
0.4
1.0
88.4
15.0
14.7
58.7
111
Emission
Period
(days/yr)

180
180
180

150, 1801
150, 180
150, 180
150, 180

150
100
100

Annual
Emissions
(Tg CH4/yr)
66
34
31
1.7
5
1.6
1.5
0.3
2.1
38
20
14
4
109
1 Temperate emissions are calculated using a period of 1 50 days, except for
temperate wetlands between 20 N and 30 N, for which a period of 1 80 days is
used.
opposed to wetland area or emission period can be a difficult and somewhat arbitrary task
because these parameters are highly interrelated.  However, some generalizations can  be
made about the most important environmental variables.

             Variables Affecting Emission Rate

                  Precipitation as it affects water level.
                  Plant community as characterized by net primary productivity, primary
                   mode of methane transport, and nutrient availability.
                  Temperature.
                                                                            Page 2-14

-------
Exhibit 2-8
Comparison of Global Emissions Studies
Study
Matthews & Fung 1 987
Aselmann & Crutzen 1 989
Bartlett et al. 1990
Fung et al. 1991
This report*
Climate Zone
Northern
(above 50N)
65
25
39
35t
34
Temperate
(20-50N;
30-50S)
14
12
17

5
Tropical
(20N-30S)
32
43
55
80*
66
Global
Estimate
(Tg/yr)
111
80
111
115
105
t Bogs and tundra (Fung et al. classified wetlands not by latitude but by four ecosystems.
However, bogs and tundra exist almost exclusively in the north, while swamps and alluvial
formations are found almost only in the tropics.)
t Swamps and alluvial formations
'Northern wetlands are defined in this report as those wetlands at or above 45 N; well-
drained tundra not included in this comparison because it was not included in any of the other
studies.
             Variables Affecting Wetland Area and Emission Period
                   Precipitation
                   Temperature
                   Actual Evapotranspiration
                   Human impacts
                   Sea level change
                   Permafrost
      A discussion of these variables follows.

2.4.1 Emission Rate

      To characterize the importance of the environmental variables that could influence
changes in methane emission fluxes, workshop  participants developed  rankings  of the
importance of the different environmental variables for major wetland ecosystems (see Exhibit
2-9).  For this exercise, the northern and  tropical ecosystems were divided into somewhat
different subgroups than those used to estimate current emissions (listed above). Northern
and tropical wetlands were divided into forested, non-forested, and open water systems.  In
addition, non-forested ecosystems were sub-divided  into graminoid  (grassy) and bryophyte
                                                                            Page 2-15

-------
(mossy) dominated wetlands. These subgroups preserve important information on available
nutrient levels, plant productivity, and plant ability to transport gas.

2.4.1.1      Precipitation or Water Level

       In those regions where precipitation is expected to increase due to climate change,
methane emissions are also expected to increase. This is because the availability of water is
the primary condition which must be fulfilled to have a wetland (Harriss et al. 1982).  Once
a soil is wet, other environmental variables such as temperature are important in determining
the rate at which soil CH4 is produced and released. In addition to being a prerequisite for
methane generation, the water level of a wetland has also been correlated with the emission
rate (Harriss et al. 1988; Bartlett et al. 1989).  As illustrated in Exhibit 2-9, the significance
of water level of a wetland can be divided into three aspects:

       Timing and Duration of Saturation. Changes in precipitation could alter the timing and
       duration of saturation of wetlands.  This timing is important in all ecosystems except
       open-water systems, which are permanently inundated.  For  example, the average
       emission  rate will decrease if  the saturation period shifts from  warmer  to colder
       months or decreases in length.

       Flushing Rate. Increases in precipitation could also act to increase the flushing rates
       of systems.   In general, higher flushing  rates result in lower fluxes.  This  aspect is
       important to alluvial systems, which currently have lower emission rates than other
       wetlands for this reason.

       Water Table Depth.   Increases  in precipitation could increase the depth of the water
       table in all wetland areas. This effect could result in higher  methane fluxes, since the
       depth of the water table has been positively correlated with methane fluxes in some
       regions.

       Information is available that describes the effects of changes in water level on methane
fluxes.  In northern  wetlands, several  researchers have determined statistical relationships.
For example, Dise (1991) relates water level, seasonal soil  temperature, and methane flux for
a series of bogs and fens that cover a gradient in wetness in northern Minnesota. The water
table explained 62% of the variability in the methane flux.  Ninety  percent of the annual
variability was  accounted for when annual soil temperature was added to the regression
equation. An additional variable that explains some  of the residual uncertainty in this model
is the Von Post peat humification index of the sites, a measure of  the degree of decomposi-
tion. An earlier effort by Svensson (1976) also reports a quantitative relationship between
methane flux and soil moisture (as %  dry weight).  Furthermore, Moore et al. (1990)  found
that although flux was poorly correlated with water table and temperature at any one of their
sites, examination of the entire data set indicated a relationship between these variables and
flux.

       The relationship between water level and methane flux is more difficult to discern in
the tropics.  This difficulty arises largely from the high variability in tropical emission  rates.
While emission measurements taken during low water generally appear to be lower than those
collected during higher water levels, it is not currently clear if they are statistically different.
For  example, the seasonal data set (averaged over the entire season) from lakes sampled
                                                                             Page 2-16

-------


co
O

co

CD
to
CO
o 
o .
0) < -a
+ CO <
* co .0
C O Q.
_O CO CO
(A . j . ,
W +* +- '
E c. c
HI II
o | z
C "o 4-*
O o C
u) '  ' W
O) ID to t!
N S E R
ti .  
r5 w -v. ""
g >  ^
il  8 ^
*- CD ^
C M
eu c "
E  in
c - ..
2 5 c
'5 o co
c *- t

M- ^ a.

 is
4> CO C
 i ..
m ^
CD
CD
co
CO
O)
c
co
oc



-
o
1

CO








*~




in



*-




in





CO





plant transport


| Exchange

-



CO




CO




^>



in




CO





in





diffusion


|| Mechanis

in



in




in




in



^










,_





O)
25
.a
D
JD




~



-




^~




in



in




in





in








1 Nutrients

Z


^
z



^f
z




in



in




in





in





CO
^o
.c
O)
JP.

CD
3
4-"
co
CD
a.
1

CO



CO




CO




in



in




in





in




CL
active season tem



Page 2-17

-------
around Manaus is indistinguishable from most of the data collected over shorter time periods
(Devol et al. 1990).  In more permanently inundated wetlands in Panama, wet season fluxes
ranged from 40% to nearly an order of magnitude greater than those during the dry season,
but were usually not statistically different due to high variability (standard errors averaged
76% of the mean in the dry season and 32% in the wet season) (Keller 1990).

2.4.1.2      Temperature

      Increases in soil temperature are likely to have the greatest influence on methane fluxes
for systems with sufficient moisture available. Two important aspects of soil temperature are:

            Length of the thaw period.
            Temperature of the season in which most of the methane is emitted (the active
             season).

      For northern wetlands, both the length of the thaw period and the temperature of the
active season can be expected to increase as atmospheric temperatures increase.  Methane
fluxes would be expected to increase, in turn, as a growing body of information indicates that
there is a positive relationship between  methane flux and these factors.   For example,
laboratory experiments have demonstrated that methane flux increases exponentially with
temperature (Koyama 1963; Moore et al. 1990; Crill, pers. comm.). Based on both laboratory
and field data, several exponential relationships have been postulated, including the following
formula from Fung  et al. (1991):

                            flux(t) = flux(t0) *  Q10(T-To)/1

      where              T    =  temperature,
                          t    =  time, and
                          Q10  =  2 (a constant coefficient; roughly equals the factor by
                                  which emissions will increase for a 10C increase in
                                  temperature).

The value of 2 for Q10 is based on observational  data from northern wetlands which would
imply a Q10 value greater than 2. However, these data have been adjusted to account for the
effects on emissions from changes in hydrology and substrate quality (Fung et al.  1991).

       The active season temperature has been recently  correlated  with emissions from
several northern habitat types, including wetlands in Minnesota, Alaska, and Sweden (see
Exhibit 2-10). A regression of this newly available data (Exhibit 2-11) shows that despite
considerable variability within each data set, the  data tend to fall along a similar line.  This
similarity is true for habitats within a single region and across arctic and boreal zones and may
imply a functional relationship between temperature and methane release across a variety of
high-latitude environments.  However, the similarity could also be a function of a limited data
base.

       The data in Exhibits 2-10 and 2-11 imply an average Q10 value of about 8 (range 5 to
50), which is a more pronounced temperature effect than postulated by Fung et. al. (1991)
(see above). However, like the data used in Fung et al. (1991), there is reason to believe that
the emission increases observed in these studies are influenced by factors other than
                                                                            Page 2-18

-------
                                   Exhibit 2-10

               Emissions Rate vs Soil Temperature in Northern Wetlands
      1000  -.
   CD
   X
   ID
   X
   o
       100 ,
         10  ,
 MN-FEN
A MK-OPEN BOG
 MN-FORESTED BOG
A AK-WET MEADOW
O AK-UPLAND TUNDRA
O SWEDEN-OMBRO #4
  StYEDEN-MINERO /
     OMBRO. #5
V SWEDEN-MINERO. #6
                                                                      V
                                  A A.

                             D

                             O
f7
 \
                                      Dl
                                                         v
                                                       D
                                                      O
                                       n   o
                      5                  10                 15

                         SOIL  TEMPERATURE  (C)
  Sources: Svensson and Rosswall (1984); Crill et al. (1992); Bartlett et al. (1992);
  Livingston and Morrissey (1992).
temperature (e.g. moisture and organic matter availability), and therefore the Q10 values may
be artificially high. High field related Q10 values may also result from a natural seasonal
increase in methanogenic bacteria. Because many of the northern latitude environments are
initially frozen sediments and methane emission rates at the onset of the experimental season
are very low values, it is probable that significant increases in the methanogenic population
size occur over the course of the experimental season (Lee Mulkey pers. comm. 1993).

      It is more difficult to assess the effects of increased temperatures on tropical wetlands.
Temperatures vary annually  by only a few  degrees in the tropics,  and tropical emission
measurements are intrinsically highly variable, making it difficult to distinguish correlations
between the active season temperature and methane fluxes.
                                                                           Page 2-19

-------
                                   Exhibit 2-11

                  Regression of Emission Rate Data in Exhibit 2-10
         1000
          100

      O)
     

      x     10
      13
      X
      O
           0.1
                 6      8

                SOIL TEMPERATURE
                                            10     12     14     16      18
                                                     o
2.4.1.3
Plant Community
      Plant populations can be modified as a result of climate change. These modifications
can  result  from  changes  in temperatures, or  precipitation,  or  from  enhanced  CO2
concentrations.   Modifications in  plant populations can be expected  to  affect methane
emission rates, since plants have a variety of complex effects on both total emissions of
methane and the mechanisms by which methane is released.  Some key characteristics of
plant communities relevant to methane emissions include net primary  productivity (NPP),
nutrients, and exchange mechanisms (ways that CH4 is transported from areas where it is
produced to the  atmosphere)  (see Exhibit 2-9)  (Whiting et al.  1991; Schutz et al. 1991;
Chanton and Dacey  1991).  The ways  in  which these  plant characteristics affect methane
emissions are discussed below.
      Net Primary Productivity.  Plants serve as the sources of organic substrate for the
      decomposition process that leads to methane emissions from wetlands. One measure
      of the availability of substrate is the net primary productivity (NPP). NPP is the gross
      amount of carbon fixed by plants minus the carbon consumed for plant maintenance.
                                                                          Page 2-20

-------
This productivity is manifested  in biomass production  and root releases.  Plant
productivity is itself controlled by other environmental variables, including temperature,
precipitation, and atmospheric concentrations of carbon  dioxide and ozone.  Thus,
methane emissions from wetlands will be indirectly affected by climate change through
its effect on plant productivity.

      With the exception of open water areas, where few rooted plants exist and
organic sources are either advected in or fall through the water column, NPP is critical
to methane emission rates. NPP can affect methane production in two ways:

      Carbon  Uptake and Root Releases. The terms carbon uptake and root releases
      refer to the carbon that is  released from  plants and is easily and quickly (i.e.,
      within a season) decomposed.   For example, non-forested  swamps in the
      tropics have higher emission rates than flooded forests in part because carbon
      from graminoid (grassy) plants is more easily decomposed than from woody
      plants.  Wetlands release more methane  during the times of year when plants
      are releasing more carbon  (Wilson et al.  1989).

      Litter Quality and Carbon Storage. The terms litter quality and carbon storage
      refer to  the quantity and quality of the carbon from plants that is dropped onto
      the soil  surface, and that therefore, may be stored for several seasons on the
      soil  surface or within  the soil.   In the tropics, where generally little peat
      accumulates, litter quality  is probably relatively unimportant (Exhibit 2-9).

Exchange Mechanisms.  There are three exchange mechanisms by  which methane
produced  in the soil layer can be transported  to the atmosphere:  plant transport,
molecular diffusion through soil and water, and bubbling. Plant transport and bubbling
are much faster mechanisms than  molecular diffusion, and the degree to which these
two mechanisms are available largely determines how much of the methane will reach
the atmosphere and how much will be oxidized.  The predominance of one mechanism
varies with latitude and habitat (Exhibit 2-9), and could change as a changing climate
alters the plant community.

      Bubbling. Bubbling is frequently the most significant pathway for  CH4 loss to
      the atmosphere in  many habitats, except for vegetated northern wetlands,
      where it has seldom been observed.   In the Amazon, for  example, direct
      measurements of bubble emissions determined that bubbles contributed from
      20% to 80% of total emissions, with lower percentages from areas of open
      water and higher figures from mats of floating grasses (Bartlett et al.  1988;
      Crill et al. 1988; Bartlett et al. 1990; Devol et al.  1990).  It is not known, at
      this point, if the occurrence of bubbling in an area merely affects the timing of
      methane release or if it  can alter the magnitude of the annual emission. The
      environmental controls  on bubbling are also not  well  understood,  but it is
      possible that climate change could increase or decrease the magnitude or areal
      extent of bubbling,  and therefore  affect global methane emissions.

      Diffusion.  This mechanism is universally available and is the most important
      one in those ecosystems where the other two mechanisms are not available or
                                                                     Page 2-21

-------
             not particularly functional, such as well-drained tundra.  However, diffusion is
             not likely to change significantly with future climate change.

             Plant Transport. The ability of plants to act as gas conduits to the atmosphere
             varies significantly with species and  plant structure, so the relative importance
             of this loss mechanism can be highly site-specific and highly dependent on the
             plant community (Sebacher et al. 1985; Chanton and Dacey  1991; Chanton et
             al. 1992).  This is true, for example, in vegetated northern wetlands, where
             bryophyte plants (mosses) have no root structures and thus do not take up and
             release CH4.   Alternatively,  graminoid  (grassy)  plant species  and aquatic
             vegetation in open water areas are more effective as conduits for release than
             are the woody structures  of trees and shrubs. Therefore,  if climate change
             affects plant productivity or species mix, global emissions could be significantly
             altered.

       Nutrients.  Nutrient levels are generally of secondary importance because they help
       determine the plant species in an area, which in turn determine NPP and exchange
       mechanisms.  While there are usually sufficient nutrients in the tropical regions to
       support the existing  plant communities,  many northern  wetland systems are highly
       ombrotrophic (fed by rainfall, not ground water) and poor in  nutrients.  This nutrient
       deficiency could limit changes in plant communities in  these regions since species
       requiring high nutrient levels  would be  unable to enter the region.  Alternatively,
       species adapted to low nutrients  could become established.

2.4.2  Wetland Area and Emission Period

  Variations in climate are likely to result in changes in both the  areal extent of wetlands and
in their annual active period of emissions.  Exhibit 2-12 provides an assessment of the relative
importance  of potential  changes in the environmental variables controlling the  area and
emission period of wetlands. The relative effects of environmental variables are the same for
most northern and tropical  wetlands.19  Although technically  not  a natural environmental
variable, human impacts are included here because they could be a very large determining
factor in the areal extent of future wetlands in some regions. The  important variables that
could affect  wetland area and the emission  period for northern and tropical wetlands are
discussed below.

2.4.2.1       Northern Wetlands

The most important variables affecting wetland area and the emission period are temperature,
precipitation, actual evapotranspiration (AET), and the depth to permafrost.20
   19 The majority of northern wetlands are non-riverine (not associated with a water body), and therefore the
relative effects discussed here are ranked with this type of wetland ecosystem in mind. The relative effects of
environmental variables on riverine northern wetlands may be slightly different than for non-riverine wetlands.
Tropical wetlands, on the other hand, are predominantly riverine, and the relative effects shown here are ranked
with this in mind.

   20 AET, or actual evapotranspiration, is defined as evaporation from soil plus transpiration from plants.
                                                                                Page 2-22

-------
Exhibit 2-12
Effects of Environmental Variables on Wetland Area and Emission Period
VARIABLE
Precipitation
Temperature
Actual Evapotranspiration
Human Impacts
Sea Level Change
Depth to Permafrost
NORTHERN
AREA
4
2
4
3
1
4
PERIOD
3
5
3
2
1
1
TROPICAL
AREA
4
0
3
5
2
0
PERIOD
5
1
3
3
1
0
* Rankings are relative and refer to within ecosystems (columns), not across
ecosystems (rows). (0 = not important; 5 = very important)
       Wetland Area.  For northern  wetlands,  changes in precipitation will be critical in
       determining the areal extent of a wetland.  However, AET and the presence of and
       depth to permafrost are also important (Hinzman and Kane 1992).  For example, if
       permafrost near the land surface melts or recedes downward, the water can drain from
       the soil, destroying the wetlands.  Similarly, increased AET could dry out wetlands.
       However, it is also possible that permafrost melting may enhance inundation in some
       areas (depending on topography), due to the collapse of the fragile wetland structure
       to the new level of the water table, once the water level has receded.

       Emission Period. Since temperature provides the major seasonal signal in the far north,
       changes in temperature will likely be the most critical variable determining the length
       of the active season in northern wetlands (Exhibit 2-12). However, seasonal changes
       in precipitation and AET are also important, since the balance between summer and
       winter precipitation is important.  If AET is higher than precipitation in the summer,
       then the emission  period may become  shorter.  Alternatively, if precipitation only
       increases in the winter, there may be no change in the emission period, since most of
       the winter precipitation falls on a frozen wetland and is lost as spring runoff.
2.4.2.2
Tropical Wetlands
       In tropical regions, precipitation largely controls the presence of seasonal wetlands and
provides the major seasonal change.  Even in more permanent systems, precipitation is critical
to both the areal extent and emission  periods of the wetlands (Exhibit 2-12).   Tropical
precipitation patterns are most often regional in scale and driven by large-scale atmospheric
dynamics, such as the monsoons. Relatively small alterations in these atmospheric features
(e.g., timing, location, frequency, intensity) could have significant impacts.  Changes in
temperature are unlikely to create large changes in areas or emission periods from wetlands
because temperatures are high and fairly constant year-round.
                                                                            Page 2-23

-------
      Sea level rise could decrease methane emissions by reducing the area of wetlands in
coastal locations.  Because many tropical wetlands are located on relatively flat alluvial areas,
rising sea level may have an impact on wetland area, type, and distribution, through simple
flooding.  Also, as sea level rises, and  especially if precipitation decreases, the salinity of
ground water may increase. A considerable data set on emissions from saline marshes across
a relatively wide latitudinal range has been accumulated which clearly demonstrates that
annual  flux is strongly and negatively correlated with  average soil  salinity (DeLaune et al.
1983; Bartlett et al. 1987).  In fact, methane release from most saline areas appears to be
relatively minimal  {Bartlett et al. 1985).  Therefore, sea level rise is considered as a variable
affecting area rather than emission rate because a wetland ceases to emit  methane when it
becomes salinated.

      Human activities,  such  as  agricultural  and  industrial development  and  water
management programs,  may have a large impact on wetlands in the future (Exhibit 2-12).
Although most of these changes will affect the area of wetlands, there may also be significant
changes in emission  periods.  For example,  water control structures will  alter patterns of
seasonal inundation.

      Most human development of wetlands is highly destructive of the original ecosystem
area; however, water management and control practices may actually create additional source
areas of CH4.  For example, tropical reservoirs in flood plains tend to be very shallow and
subject  to rapid  siltation and they therefore have  relatively short useful lifetimes.  They
become large shallow-lake systems with abundant organic material and potentially large CH4
sources. In south Florida, for example, CH4 emitted from the large water conservation areas
created for agriculture effectively balanced the loss of CH4  from  natural  wetlands due to
drainage (Harriss  et al.  1988).   In addition,  while most wetland agricultural development
involves drainage, rice agriculture often involves replacing a natural wetland with a managed,
anthropogenic wetland that produces even higher CH4 emissions. Also, wetlands within
riverine  systems are linked by the river and human development to the entire system and are
affected by activities elsewhere along  the river, including the input of nutrients, organic
materials, and pollutants, changes in sedimentation and water flow, and  changes in plant
species. All of these variables can significantly affect wetland characteristics and rates of
CH4 production and release.
2.5    Possible Future Scenarios for Methane Emissions

       In order to develop possible future  emission  scenarios of methane from  natural
wetlands, it is  necessary to outline some of the potential climatic changes over the next
century.  This section provides a brief discussion of changes in precipitation and temperature
that are suggested by available climate models.  It then develops some possible scenarios of
changes  in methane emissions as a result of these climatic changes.

2.5.1  Predicted Climate Change

       The Intergovernmental Panel  on  Climate Change (IPCC) predicts that atmospheric
concentrations of C02 will effectively double between 1990 and 2025-2050 (IPCC 1990).
The potential effects of this doubling were examined through large-scale atmospheric General
Circulation Models (GCMs). Models summarized by the IPCC yield estimates of global mean
                                                                            Page 2-24

-------
warming that range from 1.9 to 5.2C with a doubling of CO2 concentrations, and increases
in global precipitation that range from 3% to 15% (Mitchell et al.  1990).  Although these
globally and annually averaged estimates suggest the likely size of changes, regional and
seasonal estimates of climate change are more important for estimates of feedback mecha-
nisms.

      While many parts of the climate system are currently not well understood and various
models show different increases in temperatures, particularly on a regional and seasonal scale,
there are several large-scale features that appear in most model predictions.  These features
are summarized in the IPCC report (IPCC 1990):

     All models predict enhanced warming at higher latitudes in late fall and winter.  In the
      more recent higher resolution models, warming over North America during the winter
      is predicted to be 4C,  increasing to 8C over northeastern North America.  Over
      Europe and northern Asia, warming is about 4C, with  some areas of much greater
      warming (e.g. eastern Siberia).

     According to most models, warming over northern mid-latitude continents in summer
      is greater than the global mean. Summer warming in more recent models is typically
      4 to 5C over the St. Lawrence-Great Lakes region of North America and 5 to 6C
      over central Asia.

     Tropical  warming is predicted to be less than the global mean and to have  little
      seasonal variation.  Typical estimates are  2-3C.

     All  models indicate enhanced  precipitation in  the high  latitudes and the tropics
      throughout the year, and in the mid-latitudes in the winter. For example, higher
      resolution models predict an increase of 10%-20% in precipitation averaged over land
      between 35-55  N.

     All  models indicate a general increase in  soil moisture  in the northern high-latitude
      continents in the winter.  Although models  testing  the  effects of  enhanced  CO2
      suggest that both evaporation and precipitation will increase,  rates of evaporation may
      be higher than precipitation, so that soils may actually experience greater drying.

     Most models predict an enhanced drying of surface soils in the northern mid-latitudes
      during the summer.  In  higher resolution models, soil moisture  averaged over land
      between 35-55 N decreased 17%-23%.

      Exhibit 2-13  summarizes predicted temperature changes in arctic and boreal regions
for  several GCMs under  a  doubled  C02 environment.   In these Northern regions, the
temperature may be expected to rise on the order of 4 to 8C.

2.5.2 Developing Future Estimates of Emissions

      The possibility of interactive feedbacks between natural sources of CH4and a changing
climate has been recognized for some time (e.g., Hameed and Cess 1983). Clearly, the strong
dependence of emissions on environmental variables, such as soil moisture, temperature, and
organic supply,  indicates that climate change will alter emission rates. Although a variety of
                                                                            Page 2-25

-------
Exhibit 2-13
Predicted Temperature Change (50-70 N latitude) (units: C)
MODEL*
GISS
GFDL
NCAR
UKMO
SUMMER
(June, July, August)
2-4
4-8
0-4
5 - 6
WINTER
(Dec., Jan., Feb.)
5 - 12
6- 15
6- 10
8 - 10
ANNUAL
4.2
4.0
4.0
5.2
*GISS = Goddard Institute of Space Sciences, New York
GFDL = Geophysical Fluid Dynamics Laboratory, Princeton
NCAR = National Center for Atmospheric Research, Boulder
UKMO = United Kingdom Meteorological Office, Bracknell
Source: Mitchell 1989.
feedback "loops" have been  suggested,  namely between emissions from wetlands  and
elevated air temperatures (Hameed and Cess 1983), or between emissions and increased plant
productivity  due to higher C02 levels (Guthrie  1986), the magnitude  of these  effects
continues to be highly uncertain.

      While the best estimates of future emissions may in time result from process-based
models that incorporate the major physical processes of hydrology and soil temperature
dynamics, these efforts are not sufficiently advanced to provide quantitative  estimates.
Alternatively, a set of scenarios can be developed from expert judgement about the important
processes governing emissions from the different wetland systems.  These scenarios are
discussed below, after a brief outline of current modeling efforts.

      Process-Based Modeling

      Modeling has been attempted for a few wetland systems and has yielded qualitative
results.  The  primary physical components of these models are a  hydrologic and a  thermal
model. While these types of models are common to the fields of agronomy, hydrology, and
soil physics, they are not as developed for application to wetlands.  Currently, only the direct
climate  effects of changing  precipitation  and temperature can  be included in wetlands
simulation models.  Indirect  effects  of climate  change on methane flux, such as plant
community changes, have  not been included.  A limiting factor  for these models is  their
requirement for substantial methane flux data over a range of climatic conditions so that the
models can be calibrated based on actual measurements.

      Model simulations have been performed for both northern  and temperate wetlands.
Roulet et al.  (1992b) report  that modeling of northern non-forested  fens indicated  that
changes in moisture regime had significantly greater impacts than changes in temperature on
emissions. In addition, the precipitation effects varied with fen type, since some fens had
floating ground surfaces.
                                                                            Page 2-26

-------
      Harriss et al. (in press) examined the global response of CH4 emissions in northern
areas to temperature trends in historical data. The warmest and coldest years at single sites
were calculated to produce large flux differences (warm years produced emissions more than
50% greater  than emissions in  cold years).  However,  when the model was expanded to
include major northern wetland areas, asynchrony in variations in global temperatures resulted
in much smaller annual flux variation.  This model emphasizes the fact that an integrated,
global perspective must be used to assess possible responses of wetlands to climate change.
It also indicates that if temperatures increase in the future (deviating from the  historical
record), emissions could be expected to increase substantially.

      Models have also been designed for temperate systems.  For example, Pulliam and
Meyer (in press) developed an empirical simulation model for swamps on the Ogeechee River
floodplain in Georgia. The model was applied to historical climate and river hydrograph data
and to simulated altered climates. Annual emissions were strongly linked to changes in flood
plain inundation, and therefore, to river discharge.  Future emissions were thus very sensitive
to changes in precipitation, with results dependent upon the assumptions made about  the
response of evapotranspiration to  elevated temperatures. These authors found that these
hydrologic changes were more important than direct temperature effects on methane flux.

      The  EPA Office of Research and Development's Environmental Research Laboratory
(ERL)/Athens intends to perform research over the next few years to provide estimates of
future methane emissions from natural sources that are based on reproducible field data and
methods.

      Possible Future Scenarios

      Several scenarios of future methane emissions have been developed based on limited
empirical data and expert judgement of the important processes governing emissions from the
different wetlands systems. These scenarios all focus on estimating potential increases in
emissions from the northern wetlands.  Scenarios have not  been developed for tropical
wetlands because a number of factors make predictions difficult in tropical  regions.   For
example, these systems

                  May be limited in carbon (little or no peat accumulation)

                  Have not exhibited a relationship between increased temperatures and
                   higher emission rates

                  Are largely riverine, so that greater precipitation (remote and local) may
                   affect inundation and wetland area in ways that are difficult to predict.

      However, scientists have not ruled  out the possibility that methane  emissions from
tropical  wetlands could be significantly altered by climate change in  the future. Changes in
a number of variables, including  precipitation, plant productivity, and human land use, could
substantially  increase or  decrease methane  emissions  from tropical  wetlands.   To date,
however, experts considered the situation in the tropics too complex and uncertain to make
specific predictions.
                                                                            Page 2-27

-------
      Northern wetlands, on the other hand, can have large buildups of peat, and have large
areas of marginally inundated  wetlands that could be converted to much larger methane-
producing areas. Furthermore, emissions in northern wetlands have been demonstrated to
increase  exponentially with temperature,  Therefore,  it is  possible to make reasonable
predictions from this region based on assumptions about temperature and water level.

      The scenarios that have  been developed are simplistic; they focus on changes  in
emissions from existing wetland  areas and do not account for the expansion  or contraction
of wetland areas.   They also do not account  for the impact of a number of potentially
important variables, such as sea level rise, and the impact of elevated carbon dioxide and
ozone levels on plant productivity.21   Scenarios for  the northern wetland systems are
presented below.

      Lashof Scenario

      Recently, Lashof (1989)  looked  at potential increases in  methane emissions from
northern  wetlands as part of  a  larger effort to quantify  climate  feedbacks.  A  variety  of
simplifying assumptions were used to calculate the impacts of CH4 emissions. For example,
wetland area and type were assumed not to change over the  100-year time frame or longer.
In addition, changes in the water balance were omitted, due to the uncertainty in the GCMs.

      The changes in flux from northern  wetlands were estimated as a function of the
change in emission season length and the effect of increased temperatures. The function used
to estimate the temperature effect was the same one suggested by Fung et al. (1991):

          flux(t) = flux(t0) * Q10(T'To)/1, where T = temperature and t = time.

      Based on several studies with a  large range of estimates for the constant coefficient
Q10 (1.4 to 20), a conservative  Q10 range of 1  to 4 was chosen for this scenario, with  3
being the best estimate. This conservative range was selected because "it seems unlikely that
the higher values reported in some of the literature would apply to annual average emission
rates.   Very high  Q10  values  probably reflect either  the low-temperature start-up  of
methanogenesis (Q-i0 approaches  infinity as the soil temperature approaches the freezing
point) or seasonal variations involving covariance between temperature and other controlling
variables (e.g. moisture and organic matter availability)."

      Future temperatures for a world with doubled CO2 were obtained from the GISS GCM
(average summer temperatures in the high northern latitudes are estimated to increase 2 to
4C). Based on this series of assumptions and a current northern wetland flux of 35 Tg/year,
Lashof calculated that CH4 emissions from wetlands would  increase by 17  to 63 Tg/year,
with a best estimate of 51 Tg/year, for a total increase of 150 percent.
   21 A doubling of atmospheric carbon dioxide could substantially increase plant productivity and methane
emissions (Guthrie 1986). It has been suggested that, by itself, a doubling of C02 in wetlands could result in 30%
more plant matter, 40% more soil organic matter, 50% more methane-producing microorganisms, and 50% to
100% larger methane emissions (Lamborg and Hardy 1983; Marvin Lamborg, pers. comm. 1992).
                                                                             Page 2-28

-------
      Oak Ridge Scenario

      A workshop conducted by the Oak Ridge  National Laboratory on April 4-6, 1988
estimated the upper bound of the increased net flux of CH4 between northern ecosystems and
the atmosphere under projected climate change conditions (Post 1990).  The following
assumptions were used:

          A doubling in the atmospheric concentration of CO2 will cause temperatures to
            rise  5C in the boreal  and arctic  regions.  (This  was  believed to be  a
            conservative estimate, particularly for the arctic region.)

          Although the participants acknowledged that expansion of  wetlands may
            occur, for  purposes of simplicity, they  assumed  wetland areas remained
            constant.

          Changes in methane emissions from  currently dry tundra were not included in
            this estimate.

          Methane emissions due to the decomposition of organic matter released from
            melting permafrost were not  included. The possibility that methane trapped in
            permafrost could be released by melting was also not accounted for.

          Although the assumptions used to generate emission estimates were thought
            to emphasize the most significant processes controlling carbon dynamics in
            northern environments, a variety of  important processes, such as the direct
            effects  of  atmospheric  CO2 levels, species  composition  changes, and
            ecological, structural and organizational changes, could not be included in the
            calculations.
      Two climate scenarios were examined: a 5C warmer climate with unchanged water
tables (warm/wet), and a 5 warmer climate with water tables lowered by 10 cm (warm/dry).
Methane emissions from northern ecosystems responded in roughly the same manner to both
scenarios. A calculation was not made for a warmer/wetter climate, or one with increased
water tables, because it was assumed that such a climate would respond in the same manner
as the warm/wet scenario.

      Northern ecosystem responses were separated into two systems. Predictions were
calculated in a different manner for each system:

      Arctic Response.  Methane emission estimates in the arctic region were based on the
      assumption that permafrost melting and increased drainage, regardless of precipitation
      changes, would lower the water table, thus stimulating peat oxidation. Peat would be
      oxidized at a rate of about 1 cm of peat depth per year, over an area of 2x106 km2.
      (The carbon content of 1 mm of peat is assumed to be 65 g C/m2.) While most of the
      increased soil respiration would be aerobic, or CO2-producing, about 10% to 20%
      would be anaerobic, or CH4-producing.  Methane emissions from arctic ecosystems
      were predicted to increase by at least 130 Tg/yr, with the range probably closer to
      170-340 Tg/yr (see Exhibit 2-14).
                                                                          Page 2-29

-------
      Boreal Response. For the boreal region, the average current microbial respiration rate
      is 267 mg CH4/m2/day. (It is not clear from the workshop report how much, if any, of
      this generated methane is oxidized, and how  much is emitted.)  Methane respiration
      in the  boreal region is expected  to  increase  by 10%  with every  1 C rise in
      temperature.  The  increase in methane respiration will be  entirely emitted to the
      atmosphere, while the amount of methane oxidized will remain constant. Boreal peat
      lands were predicted to emit an additional 32 Tg CH4 per degree Celsius of temperature
      rise.  Thus, for a 5C temperature rise, boreal emissions were predicted to increase
      160 Tg/yr. Notice that this conclusion assumes a linear relationship between methane
      emissions and temperature rise, as opposed to the logarithmic or exponential response
      postulated in Section 2.4.1 of this report.

      The estimates of potential increases in methane emissions  are fairly large.  These
calculations are supported  by the fact that the predicted rise in  global methane release is
similar to the observed methane rise that took place when comparable changes in climatic
regimes occurred at the end of the last glacial period (Post 1990).
Exhibit 2-14
Oak Ridge Workshop Estimates of Increased Northern Ecosystem Methane Emissions
(Tg CH4/yr)
Ecosystem
Arctic (tundra)
Boreal (peatland)
Total
Warm/Wet Scenario
130
160
290
Warm/Drier Scenario
130
160
290
       UNH Workshop Scenarios

       Participants at the 1992 workshop at the  University of New Hampshire (UNH) also
attempted to assess the potential increase in the magnitude of methane emissions under
possible future climate regimes.22  Estimates are based on current CH4 emissions of 35
Tg/year from the arctic and boreal zones (Bartlett et al. 1990; Fung et al. 1991; this report).
Three different scenarios were developed to represent a range of possible future climates.
They are all based on the assumption that atmospheric temperature will increase 4  to 6C
by the end of the next century. The scenarios differ only in their assumptions of changes in
average soil moisture.
   22 In order to assess the potential impacts of climate change on natural wetlands, a workshop was held at the
 University of New Hampshire on March 10 and 11, 1992. A list of the wetlands researchers who participated in
 this workshop is provided in Appendix A.


                                                                             Page 2-30

-------
             Scenario 1 - Warmer Temperature and Constant Water Table Scenario

      The first scenario assumes that temperatures will increase by about 4 to 6C by the
end of the next century and that the level of inundation of the northern wetlands remains
unchanged.  It was agreed that a distinct, positive relationship exists between emissions and
temperature along the lines  of the Q10 model.  However, there was general skepticism that
the temperature effect alone would  be as pronounced as it is in Exhibits 2-10 and 2-11  (i.e.
Q10  = 8).  The  results of these experiments are  believed to  include the effects of other
variables in  addition  to temperature. It was thought that a  Q10 value of  4  was a more
reasonable estimate,  given the assumptions of this  scenario. Therefore, with a temperature
rise of about 5"  C, methane emissions  would be expected to double to approximately 70
Tg/yr.

             Scenario 2 -- Warmer and Drier Scenario

      For a warmer and drier climate, assessing future emissions is more complex because
a number of CH4 feedback processes would be present. Exhibit 2-15 is a simplified schematic
showing  the major methane feedback processes  in  this climate scenario.  (Note  that gas
hydrate and permafrost methane feedbacks are not shown here.)  These processes include
both fast and slow responses. For example, increased temperatures would quickly increase
anaerobic decomposition rates and methane emissions.  A slower response such as drying of
wetlands  leads  to reduced wetland areas, thereby  increasing  aerobic  respiration  and
decreasing methane emissions.  Another slow response is the  potential collapse of the peat
structure once it has been sufficiently dried and subject to oxidation. This collapse could
create a topographically more heterogeneous landscape with CH4 "hot spots" -- low wet areas
with very high CH4 emission rates.
      Under this scenario, it is also possible that the water table  will drop significantly
throughout the entire region as air temperatures increase by 4 to 6C. In this case, emissions
may be expected to drop to roughly 20 Tg/yr, in the  short term (20-30 years), as surface soils
dry and surface peats are decomposed and degraded.  The increased  oxidation rate of peat,
however,  could lead to a collapse of peat structure  in some areas, and thus to an increase in
topographical heterogeneity and a long-term increase in methane emissions.  Emissions for
the area could potentially double to 70 Tg/year. Admittedly, this prediction is not based on
an equation  but relies on scientific judgement.  This method was chosen over the Q10 model
because  it was felt  to more accurately represent the complex, and as yet unquantified,
processes that would be involved in this scenario.

      Similar results are  obtained  by looking  more  closely at  the  northern  wetland
ecosystems. For example, the boreal lowland areas of Hudson Bay and Siberia may be the
regions that will experience the greatest changes under such a climate scenario. Assuming
that these systems currently emit about 2 to 5 Tg/year, emissions could increase to between
10 and 30 Tg/year as a result of drying, followed by the collapse of the peat structure and
the formation of  methane "hot spots." The rest of the boreal region (representing 15 to 17
Tg/year)  may produce lower emissions over  the short term,  followed by a return to their
original emissions values. In total, emissions could reach about 40 to 60 Tg/year, which is
an increase of 5 to 25 Tg/year.
                                                                           Page 2-31

-------
                                   Exhibit 2-15

               Methane Feedback Processes in a Warmer, Drier Climate
                               I ncreased GIobaI
                               farm}ng Potent I a I
         Reg Iona
     Warmerj  Drier
        Ecosystem
                                       Increased Anaerobic
                                       Decompos111 on  Rate
                            C SLOVQ
 Plant Success Ionj

deduced Wetland Areaj

Increased "Hot Spots'
                                                        Increased ChM
                                                          OxldatIon
             Scenario 3 - Warmer and Wetter Scenario

      Fast and slow feedback mechanisms also exist in a warmer and wetter scenario (as
illustrated in Exhibit 2-16).  These mechanisms include  increased emission rates (fast) and
increased wetland areas (slow).
      In this scenario,  water tables are assumed to rise and air temperatures are assumed
to rise by 4 to 6C. Under these conditions, both the  wetland area and the emission rate
are likely to increase. Because of its wide distribution and large area, well-drained tundra may
be a critical habitat in this scenario.  Increased precipitation could  convert these areas to
significantly larger CH4  sources, and emissions may rise to as high as 100 Tg/year.

      The estimates in  the three UNH scenarios are highly uncertain for a number of reasons.
First, the change in the area of wet surfaces (inundated wetlands as well as lakes and ponds)
is very tentative. This causes doubt in the emission estimates, because emissions from these
sources are quite high relative to the surrounding terrain.  Second, uncertainty arises from the
implicit assumption that northern areas from which there are no flux data are  similar to sites
already  examined.  Third,  these estimates are limited to direct climate effects and do not
include  possibly significant indirect effects, such as those due  to changes in photoperiod
                                                                           Page 2-32

-------
                                     Exhibit 2-16

               Methane Feedback Processes in a Warmer, Wetter Climate
          Regiona
      Warmer.,  Wetter
          Ecosystem
                                  I increased GI oba I
                                 Warming Potential
                                         Increased  Anaerobic
                                         Decomposition Rate
                                           Plant  Success!onj
                                        Increased Wetland  Area
Atmospherlc
CH4 Increase
                                                           Reduced  CH4
                                                           Oxldat ion
length23 or atmospheric C02 concentration.

       Results from all of the northern wetland scenarios are summarized in Exhibit 2-17.
While the range of predictions is large (5 to 290 Tg/yr), in all cases emissions are expected
to increase over the next century.

 2.6   Uncertainty and Further Research Needs

       Much of the  considerable  uncertainty surrounding  current  and future methane
emissions from natural wetlands can be reduced with additional research.  The following
section contains an estimate of the uncertainty in current emissions as well as a discussion
of research needed to reduce this uncertainty.  Section 2.6.2 addresses the  difficulty  in
estimating future emissions and discusses several types of experiments that will  make future
predictions more accurate.  Finally, Section 2.6.3 examines the  phenomenon  of methane
uptake by soils and how this contributes to uncertainty.
   23 Photoperiod is the daily period when photosynthesis is the dominant process in a plant (approximately equal
to daylight hours). Plants at very high latitudes may respond differently to climate change than those at lower
latitudes due to the difference in photoperiod during the summer months.
                                                                             Page 2-33

-------
Exhibit 2-1 7
Summary of Northern Wetland Emission Scenarios*
Author
Lashof
Oak Ridge Workshop
UNH Workshop
Warmer/Drier

290
5 -35
Warmer/Wet
17-63
290
35
Warmer/Wetter


65
* Values shown are for the change in emissions in Tg CH4/yr.
Sources: Lashof 1989; Post 1990; this report.
2.6.1 Current Emissions

      The amount of methane currently being emitted from natural wetlands is uncertain but
the range of estimates is believed to be fairly accurate.  This is due to the similarity between
the estimates for global emissions derived in this report and in several previous efforts. While
most of these efforts are based on similar data and assumptions, others used very different
methodologies (e.g., Fung et al. (1991) arrived at a similar estimate using a tracer model).

      Quantifying Uncertainty

      It is difficult to perform a rigorous statistical analysis of the uncertainty in the current
emission estimates derived in this report. An approximate attempt to estimate the upper and
lower bounds of the current emissions from each of the wetland types follows. These bounds
are based  on ranges of uncertainty estimated for (1) the emission rate and (2) wetland area.
Uncertainty  in the emission period is not included in these ranges.  The  purpose of this
exercise is not to give a definitive range for current  emissions but rather to convey the
approximate magnitude of such a range.

      Emission  Rate Range.  A common statistic used for estimating  the variability in a
      population is the standard error of the mean (see Exhibit 2-18).  The upper and lower
      bound of the emission  rate is one standard error of the mean in each direction of the
      average.   The standard  errors are considerably higher for tropical wetlands (about
      44% of the mean) than for northern wetlands (about 25% of the mean).  Exhibit 2-18
      suggests that a  major source of uncertainty in current emissions is the emission rate
      rather than the  area of emissions.  Matthews and Fung  (1987) and Aselmann and
      Crutzen (1989) also identify the emission rate as the major source of uncertainty
      (Matthews, pers. comm. 1992).

       Wetland Area Range.  The  range  of areal uncertainty is  derived not by statistical
      analysis but by comparing  small  increments of areal estimates provided by two
      separate researchers who used different methods:  Matthews and Fung (1987), and
      Aselmann and Crutzen (1989).  The areal estimates  of Matthews and Fung (1987),
      which are used as the  baseline in this report, were derived by combining three
       independent digital  data bases for  (1)  soils,  (2) vegetation, and (3)  inundation.
                                                                            Page 2-34

-------
      Aselmann and Crutzen (1989) used regional wetlands surveys and monographs to
      derive wetland areal estimates.  While these two  research teams used different
      vegetation classifications, they both give total wetland areal estimates by 10 latitude
      bands. Thus, by taking the smaller of the two estimates for the five latitude bands in
      the tropical region, a lower bound for tropical area is estimated. Similarly, lower and
      upper bounds are estimated for temperate, boreal, and arctic regions.  According to
      this method, the tropical wetland area ranges from 14 to 25x1011 m2, and the northern
      wetland area ranges from 88 to 105x10  m2.  This method for estimating the areal
      range was chosen because neither Matthews and Fung  (1987), nor Aselmann and
      Crutzen (1989) give a range for the area of wetlands.  It is entirely possible that the
      areal range derived here does not incorporate the full degree of uncertainty surrounding
      wetland areas.

      Emission Period Range. The emission periods used in this analysis are from Matthews
      and Fung  (1987), who did not estimate a range for period used. While it is generally
      agreed that the Matthews and Fung values are approximations, a reasonable method
      for estimating uncertainty in period was not found.

      Given  the uncertainty ranges for emission  rate and  wetland area, the range of
uncertainty in total wetland methane emissions is estimated to be from approximately 70 to
170 Tg/yr (see Exhibit 2-18).  This is a rough  estimate since not all  sources of uncertainty
could be quantified.  More of the uncertainty comes from the tropical region (where emissions
range from 34 to 118  Tg/yr), than from northern wetlands (29 to 52 Tg/yr).

      Research  Needs

      Uncertainty in current emissions can be reduced with more and longer term studies.
Because regional flux comparisons show that data from one area cannot be easily extrapolated
to other regions, a primary research need is data from large wetland areas from which there
are  currently little or no data. Also crucial to the understanding of methane fluxes are multi-
year flux and environmental variable studies such as the report by Whalen and Reeburgh (in
press). These types of studies need to be conducted on decadal  time scales, which offer the
only reasonable way to assess natural variability in flux and the integrated response of the
wetland ecosystem.  Long-term flux  studies also permit the examination of the complex
relationships among variables controlling the flux.  More specific data deficiencies and
research  needs are discussed below for tropical, temperate, and northern  wetlands.

             Tropical Emission Data

      Uncertainty in  current emissions  from the tropics is  greater than the uncertainty
surrounding the temperate or northern regions.  Less sampling  work has been done in the
tropics.

      Important tropical regions from which  there  are no data include: (1) large  African
wetlands, such as the Sudd, the headwaters  of the Nile River, and the Okavango, (2) the
Pantanal  region in Brazil and Bolivia, and (3) Indonesian peat swamps. On the other  hand, a
considerable and relatively consistent data set has now been  assembled for the Amazon
(n = 788 and nearly annual coverage).
                                                                            Page 2-35

-------






















1
1
m
(O
1
UJ
? 1
gO n
+* 0>
1 ?
^5 *
X e
UJ |
O
.E
1
a
0
c









^^
W3 t^
C -?*
.C O *-
o> '55 I
? w O
J-" *?
E co

*""""*

05 >
> o ~>
> -co X
5  0
E co
UJ I
* '
CO "*

a  ">
2 w X
 *p
<  03
m b

o
ir J
l|

 *.
"c
co^F
|o


C
^^ C Q^
UJ

c
5 a>
 ~ 2
UJ

2 ^
LJJ ^_
"^3
+5 ^^
CO 0
CO ~
ill
<| 0,
UJ ^


Ecosystem




00





CO






m
CD




in
CM
2




03



















Tropical
Wetlands




o
CD





2






CO




2
a
q
00




o



in
CO
CM




in
O3


o


in
CD




Flooded
1 Forests




in
in





o
CM






CO
CO




o
o
CO




00



in
00
CM




00


CM
in


CO
CO
CM



 CO
S o.
H- E
c 5
o 5
Z CO




CO





2






CO
r-




c>
o




If)
6



o
in
CM




CO'


CM
O


00




C 0)
0) 
Q 5




in





in






in




CO
CO




5


















o
I Temperate
Wetlands




CM
m





03
CM






CO
CO




in
O
CO
00




00
00



















1 Northern
Wetlands




00
CM





*






o
CM





in




in



00
o




CD
CO


00


00




"co
0




00





^






2




CD
in




m



J




00


CM


CO
O3




O




CD





CO






*




"CM
CO
in




OJ
in



en




in


CM


-




(Drained
Tundra




in





00
CO






en
0




CD
CO
00
























"55
S
0)
CO L-
a) a>
CO .
CO _,
 ^
*S ^ CO
co -^ -5
T3 E ^
"O o co
c -t 
S .2 
5 S 
03 > CO
W -S TJ
03 ^ C
~-- co o
CCO _Q
^j ^^
S c -
II 1
^"^ ^L
"o oi "2
* w C
a E 
c co ^
s 2 g
E6 t
11 |
c: o -P

_
C S ^ c Q>
nj ^_   ~ tz
1 a>- S | _:
CD  O  c 00
J? C a) c 00
r o co 53 03
Jr 2 "c g 0 .C
2 ilj  >  E
a co 5 c o a)
|5 -a ^ co "to ,2
^ c t! 1= w ^
C 5 CO 5 'c T3
 /n C
m  ">
+ "o !^ co c
' * ^ cl)
a> aj c Q. ^ -55
^ (r .<2 fc  >
 C CO CD *^
 .2 g-  e E
S S o ^E -2 2
1 E " 5 c E c
ijj uj E t: .0 -c 
g, S>t    
2 2 -o .g ^1 " co
<
-------
      Yet, while average emission rates have been well characterized for Amazonian wetland
types, the area and emission period  of these types are uncertain.  The  emission period is
determined primarily by flood waters, which annually inundate and recede from the flood -
plain. Seasonal remote sensing data of vegetation and inundation throughout the flood plain
will greatly reduce these uncertainties.

             Temperate Data

      While temperate wetlands make only a small contribution to global emissions, they are
well-characterized in terms of emission rates, particularly in the U.S.  Furthermore, extensive
ancillary data bases in the U.S.,  containing data such as air temperature, rainfall, and  river
discharge  rates have  facilitated attempts  to link patterns in  wetland emissions with
environmental variables in general. However, despite the large data set available, significant
uncertainty is associated with emissions from  this region.  The uncertainty stems from the
fact that while most studies in the temperate zone have had year-round sampling programs,
few of these studies have sampled over a period of several years.  Thus, the true, long-term
average emission rate may not be as well defined as it might seem.

      Another source of persistent uncertainty is the error caused by using measurements
from a relatively small area to represent a much larger one.  Uniformity in emissions patterns
from a small area can give the false impression that emission patterns from the larger area are
also fairly uniform. For example, working in the Florida Everglades, Bartlett et al.  (1989)
found that measurements collected within a few meters displayed significantly less variance
than those more widely separated (100 m or more). These results suggest that (1) the range
of average methane emission rates derived from the study would lead to conservative
estimates of the confidence in the means and (2) that spatial, within-system variability could
introduce significant uncertainty in extrapolations to larger scales.  This level of variability
illustrates some of the problems inherent in simple, latitudinally-averaged emission estimates.

      Small-scale variability  has also been  observed for emission periods.  Wilson et al.
(1989)  noted that wetland types within a  relatively small area in Virginia  had different
emission periods.  The  timing of  the emission period also varied among sites.

      If these well-researched temperate and subtropical sites can serve as models for those
that are more remote, variability in the emission rate and period (on scales ranging from  one
m2 to regional)  greatly increases the uncertainty in large-scale estimates and  should be
addressed in all sampling programs.

             Northern Data

      Estimates of emissions from northern wetlands are uncertain because of year-to-year
variability and winter emissions, as well as large-scale extrapolation between heterogeneous
wetland  systems (e.g., comparison of  Minnesota  and  Hudson  Bay Lowlands data).
Uncertainty in this region is greater than in the temperate region, but not as large as in the
tropics.

year-to-year  Variability. The majority of data  from the north  was collected during a single
season.  Therefore, little is known about annual variability.  Multi-year  studies have been
conducted at only two sites:  Marcell Forest (in the boreal zone) and Fairbanks (in the arctic
                                                                             Page 2-37

-------
zone).  Data from these sites suggest that emissions may vary by as much as an order of
magnitude from year to year.

Winter Emissions.  While little work has been done on this subject, emissions during the
winter months (November through March) can contribute as much as 20% of the annual flux
from northern wetlands, even under heavy snow cover (Dise, submitted (b)). The contribution
of winter emissions to annual fluxes appears to be highly variable by year and by  site. Heavy
early snow and the depth of the snow-pack which insulates surface soils may be important
controls on flux (Dise, submitted (b)). Since more significant winter fluxes were observed in
Minnesota (boreal)  than in Fairbanks, Alaska (arctic), they may occur more frequently as
winter temperatures become more moderate in the boreal region.  If winter season emissions
from the boreal zone are disproportionately  larger than those from the arctic, differences
between the two regions may be greater than they now appear.

      Much of this uncertainty, however, can be reduced with additional data  from both a
broader variety of sampling sites and from year-round, multi-year data sets. Additional survey
measurements of flux are needed in areas such as the vast Siberian lowlands because regional
comparisons  have shown that extrapolations from  one  region  to  another are unreliable.
Furthermore, recent data has indicated that moist-to-dry boreal wetlands and boreal and arctic
lakes can contribute significantly to global emissions.  However, it is not possible to calculate
the contribution from these systems at this time because the areas of these habitats are not
known.  A survey of  the areal extent of these ecosystems  is necessary. The volume of data
accumulated in only the last few years suggests that progress in decreasing uncertainties can
be rapid.

2.6.2 Future Emissions

      Estimates of potential future emissions from wetlands are much more uncertain than
current  estimates.  This greater degree of uncertainty is indicated by the large range of
predictions that have been made for future northern wetlands.  However, additional research
in several areas could help reduce the uncertainty in current emission estimates and in future
projections.

      Environmental Variables

      Only the direct climate effects of changing precipitation and temperature have been
included in the  simulation models and expert predictions.   However, several  other
environmental variables could play an important role in future emissions.  Poorly understood
effects that could significantly affect future emission scenarios include:

Solar Inputs.   Alterations  in solar inputs such as variations in cloudiness (with or without
changes in  precipitation)  could  affect  AET and plant   productivity,  and  therefore alter
emissions. Because evapotranspiration  may be a critical variable controlling inundation  in
many wetlands, an understanding of the response of AET to changed climates  is important
(Dooge  1992).

Natural Fires.  The severity and frequency of natural fires in northern latitudes could increase
as a result of  climate change, especially if wetlands become  drier.  Increased  natural fires
could have a large, positive effect on wetland methane emissions (Post 1990).
                                                                             Pag 2-38

-------
Precipitation Dynamics. The relative balance between precipitation falling in the winter, as
snow, and in the summer, as rain, is important. Since snow is accumulated on the frozen
ground surface and then released as a pulse during spring warming, its effect on soil moisture
is significantly different than that of rainfall, which is more likely to be absorbed into the soil.
Knowledge of  regional topography  and  relative elevation  is also critical  to predicting
inundation. Currently, these types of data are quite difficult to obtain for most wetland areas.

Plant Community.  Methane emissions appear to be strongly linked to the type and amount
of wetland plants present in a variety of environments (Sebacher et  al. 1985; Whiting  and
Chanton 1992).  It is possible that significant changes may occur in these plant communities,
either through  changes  in species or through  greater productivity  due to increased
temperature, moisture, or atmospheric C02 levels (Guthrie 1986; Idso 1989). This possibility
suggests that these linkages need to be better understood for more species and in more
habitats.

      Empirical Data Needs

      While the wide array of environmental variables and their  uncertain relationships to
methane emissions make  predicting  future  emissions  difficult, there  are  a  number of
observational and manipulation experiments that can help resolve uncertainty.

Multi-Year Observations.  Multi-year flux and environmental variable studies such as the report
by Whalen and  Reeburgh (1992) are crucial to the understanding  of CH4 flux. These types
of studies  need to be conducted on decadal time scales to assess natural  variability in flux and
the integrated response of the wetland ecosystem. Long-term flux studies also permit the
examination of the complex relationships among variables controlling flux.

Indicator Ecosystem Observation. It may be useful to focus some long-term research on an
"early warning"  or indicator ecosystem, which could  detect possible climate change /  flux
relationships. A useful ecosystem for this purpose would have a fairly long flux data base to
assess "natural" variability,  and would have to cover fairly large spatial scales (to ensure that
changes are widespread). The extensive areas of moist-to-dry tundra in the northern latitudes
may be a reasonable choice for these purposes since they are  "poised" between being a CH4
source or  a sink and lie within a region projected to undergo significant climate change.

Human Impacts Observation. The long-term effects of human wetland  modification on CH4
flux are largely  unknown. It is therefore important to examine emissions from  a variety of
impacted areas over long time scales and histories of plant succession. Modified areas should
include such systems as  burned  wetlands, reservoirs, areas under new water management
controls, and riverine wetlands  subjected to changed patterns  of  nutrient inputs and/or
sedimentation.  The variables that need to be considered in evaluating human impacts include
the rate of change of environmental modifications, possible interactions among climate change
and human impacts, watershed  changes,  higher CO2  levels,  increased pollution/nutrients,
plant species changes, and altered temperatures.

"Natural"  Ecosystem Manipulation.   Natural ecosystem or mesocosm manipulation of
environmental variables offers another way to examine flux relationships (Billings et al. 1983).
Such experiments typically involve  artificially altering one variable  (e.g., ambient carbon
                                                                            Page 2-39

-------
dioxide) in an otherwise unchanged natural ecosystem.  "Natural" experiments are likely to
be the most realistic and the most easily applied to the real world.

Laboratory Manipulation. Small systems or cores and laboratory manipulations may be helpful
in interpreting results from larger systems,  but in themselves are so artificial and isolated that
they are difficult to extrapolate realistically.   More  extensive evaluation  of correlations
between CH4 flux and environmental variables in models to predict responses under changed
climates, or for extrapolations into unknown areas, is needed.

2.6.3 Methane Uptake in Wetlands

      Adding to the uncertainty about the net role wetlands will play in future atmospheric
methane concentrations is  the fact that  wetlands can act as methane sinks as well as
sources. In addition, the amount of atmospheric methane absorbed  by wetlands could be
altered as the climate changes.

      In addition to oxidizing locally generated  methane before it  escapes into the
atmosphere, wetlands can oxidize or absorb atmospheric methane.  Aerobic (oxygen-present)
surface soil conditions are conducive to methane oxidation.  Therefore,  uptake of atmospheric
methane can take place when a seasonal wetland, such as a seasonally flooded forest, is dry
at the air-soil interface.  In this situation, such a wetland may be a net sink of methane. Other
non-methane-producing ecosystems are permanent sinks of methane.  The global soil sink is
currently estimated at 10 to 50  Tg CH4/yr (IPCC 1990). Similar to global methane emissions,
the global uptake of methane  is determined by the uptake rate, the uptake area, and the
uptake period.

Uptake Rate.  Since methane uptake in natural soils was first observed in 1982 (Harriss et al.
1982), this phenomenon has been extensively studied.  One of the most striking features of
the data is the  overall similarity  of uptake  rates  across ecosystems and latitudes (see
Appendix G).  Average or median uptake rates are almost always less  than -5 mg CH4/m2/d,
and usually less than -2 mg CH4/m'-/d.  Thus, while methane emission rates can vary over
several orders of magnitude ( 0 to  2000+ mg CH4/m2/d), methane  uptake rates can vary
only slightly (0 to -5 mg CH4/m2/d),

      Although CH4 oxidation in soils is a microbial process, methane uptake rates  appear
to be controlled not by biological or environmental variables, but by rates of diffusion in soils
(Steudler et al. 1989; Born et al. 1990; Crill  1991; Goreau and de Mello 1985; Born et al.
1990;  Keller et  al. 1991).  Of course,  soil diffusion  is greatly affected by inundation.
Therefore, methane oxidation rates could be affected by climate change but not to the extent
that emission rates can be altered.

Uptake Area and Period. The period and areal extent of methane uptake from non-methane-
producing ecosystems, such as upland  forests and dry savanna, are not likely to shift with
climate  change.  However, the period and area of uptake from wetlands could change if the
period and  area of methane emissions from wetlands are altered.  For example, if wetlands
become drier and the period of emissions decreases, then the period of uptake could increase.

       Future  changes in methane uptake could act as a magnifier of changes in methane
emissions:  if, for example,  a wetland location dries out and ceases to emit methane, it may
                                                                            Page 2-40

-------
become a negative emitter (uptake).  Conversely, a location that is currently a weak sink
(uptake) could become a strong source (emission) if it is sufficiently inundated.  Thus, future
predictions of methane emissions from  wetlands are made slightly more uncertain by the
magnifying effect that methane uptake could have on changes in the wetland area and the
emission period.  However, changes in global soil uptake of methane are likely to be small in
comparison with changes in global methane emissions.
                                                                            Page 2-41

-------
                                    CHAPTER 3

                            NATURAL FOSSIL SOURCES
      There are two natural fossil sources of significance: gas hydrates and permafrost.
These sources are estimated to emit about 3 to 5 Tg of methane annually, representing about
3% of emissions from natural sources and 1 % of total methane emissions. Currently, fossil
sources do not contribute significantly to atmospheric methane concentrations.  However,
emissions from fossil sources could increase by more than two orders of magnitude as a result
of climate change, making them the largest single source of methane.

      This chapter presents background information about hydrate and permafrost sources
of methane, provides estimates of current emissions  from these sources, and discusses the
potential increases in emissions as a result of climate change.

3.1   Gas Hydrates

      The current and future potential for the methane release from gas hydrates depends
on several important factors. These factors include the total quantity of methane reserves
stored in hydrates in different regions, the stability  conditions required to maintain these
reserves in storage, and the extent to which the environmental conditions of existing hydrates
could be altered in the future.

3.1.1 Background

      In order to understand the factors affecting gas hydrates, it is useful to first understand
the nature and origin of gas hydrates.

      What are Gas Hydrates?

      Methane gas hydrates are solid structures composed of rigid cages of water molecules
that enclose methane molecules.  Ordinarily, when water freezes, it forms ice in a hexagonal
crystal structure.   However, when a threshold concentration of methane or other gases is
present, and pressure is sufficiently high, water crystallizes in a cubic lattice that traps the
gas molecules. The hydrate  structure is sensitive to temperature, salt, and other impurities,
and is only stable under pressure equivalent to at least 180 m of soil overburden (Bell  1982).
The ratio of CH4 to H20 in a methane hydrate can be as high as 1 to 5.75. Because of the
high CH4:H20 ratio, a unit volume of hydrate can  contain up to about 170 times as much
mass of methane as a unit volume of pure methane gas at standard conditions (Kvenvolden
1991e).

      Methane gas hydrates are formed when methane is created within or enters into the
zone of  hydrate stability (see Exhibit 3-1). The source of methane for the formation of gas
hydrates may  be either microbial or thermal in origin.  Microbial methane is created by the
anaerobic digestion of organic matter by microorganisms in shallow, oceanic and continental
sediments. Thermogenic methane is produced by the thermal alteration of organic matter that
is deeply buried in  sediment.
                                                                             Page 3-1

-------
                                    Exhibit 3-1

                      Hydrate Stability Zone: Continental Case
                                                  Mbwnuin D0poi
                                                    Boundary
                                (io)        o        10
                                     Temperature fC)
 Source: Kvenvolden and Me Menamin (1990)
      Methane Hydrate of Microbial Origin

      As described in the chapter on natural wetlands, methane of microbial, or biologic,
origin is created by the decomposition of organic matter in wet, oxygen-depleted sediment.
In the oceanic case, the organic matter is called "marine humus", and consists primarily of the
decomposed tissue of tiny organisms such as plankton and nekton.  This humus falls to the
ocean floor, where it is buried by sediment at the rate of 10-20 cm/100 yrs (Revelle 1983).
Once it  is buried, anaerobic conditions soon prevail and the organic matter is digested by
methanogenic bacteria.  If more methane is produced than can be dissolved by the water in
the pores of the sediment (80-160 mmol CH4/L), and temperature and pressure conditions are
met, then the methane is converted into methane hydrate.

      Methane Hydrate of Thermal Origin

      Thermogenesis of methane takes place under high  pressure  O200 bars)  and
temperature (>80-120C) (MacDonald 1990). Typically, such conditions exist well below
the zone of hydrate stability.  When methane is created in this manner, it enters hydrate form
only when a migration pathway exists from the lower sediment to the upper sediment layer
(Kvenvolden  1988).  Tests  of the isotopic composition  of  the methane extracted from
hydrates at  several depths and locations worldwide reveal that most of the methane is
microbially generated, not thermally produced (MacDonald 1990).
                                                                            Page 3-2

-------
      Gas Hydrate Reserves

      The quantity of methane stored as gas  hydrates is substantial.  These methane
reserves are estimated for three major types of hydrates (see Exhibits 3-2 and 3-3):

      Oceanic Hydrates.  Oceanic hydrates are found in underwater sediments of the outer
      continental margin at all latitudes. They only occur where the ocean floor is greater
      than 300 meters below the ocean surface. They are located between the ocean floor
      and a sub-bottom depth of  about  1,100 meters, depending on the ocean-bottom
      temperature and the geothermal gradient (Kvenvolden 1991e).

      Onshore Continental Hydrates. Continental, or permafrost-associated hydrates, can
      occur within or below the permafrost zone, and are found onshore at high latitudes,
      at depths of about 200 - 1,200 m below  the  surface  of the earth in regions  of
      continuous permafrost (Kvenvolden 1991b).

      Offshore Continental Hydrates. Continental  hydrates can also be found  offshore,  on
      the nearshore continental shelf, where melting subsea permafrost has persisted since
      times of lower sea level. Since the last ice age, 18,000 years ago, sea level has risen
      about 100 - 125 m, and the temperature  of the present'shelf has risen  about 15C
      (Hill et al. 1985; Kvenvolden 1991 a; Osterkamp,  personal communication).
                                    Exhibit 3-2

                         Relative Location of Hydrate Types
                                                                            Page 3-3

-------
      The existence of gas hydrates has been verified using a number of methods such as
direct observation and seismic reflection methods. Gas hydrates are known or inferred to
exist throughout polar onshore and nearshore permafrost regions, and in the oceanic outer
continental margin sediments of all seven continents.

      Estimates of methane hydrate reserves vary widely. Oceanic methane estimates range
from 2.3x106 Tg (Mclver  1981) to 5.5x109 Tg (Dobrynin  et al. 1981), and continental
estimates range from 1x105  Tg (Meyer  1981) to 2.4x107 Tg (Dobrynin et al. 1981).
However, a number of recent,  independent estimates, based on an expanded data set, have
converged on estimated methane reserves of about 1 x107 Tg for oceanic reserves and 5x105
Tg for continental reserves  (Kvenvolden 1991 e; MacDonald 1990).  This amount of methane
is more than two thousand times the current atmospheric reservoir of methane, and  about
twenty thousand times the  current annual emissions. Of the continental hydrates, about 70%
are onshore and 30%  are offshore (Kvenvolden 1991e).
Exhibit 3-3
Methane Hydrate Reserves (Tg CH4)
Hydrate Type
Oceanic
Cont. (Onshore)
Cont. (Offshore)
Range of Estimates
2.3x106- 5.5x109
1x105 - 2.4x1 07
Best Estimate
1.0x107
3.5x105
1.5x105
       Stability of Gas Hydrates

       Gas hydrates are stored or maintained in a "stability zone." The stability zone is the
layer in the sediment where the temperature and pressure conditions are sufficient to support
hydrates.  The minimum depth of hydrate stability is proportional to  pressure and inversely
proportional to temperature (i.e., hydrates are stable if the pressure is sufficiently high and the
temperature is sufficiently low). Hydrates do not exist at the surface because, even under the
coldest conditions, the pressure is riot sufficient. Conversely, hydrates do not exist at great
depths because temperature increases with depth in sediment (according to a geothermal
gradient of 0.016 to 0.053C/meter), and by about 2 kilometers below the sediment surface,
temperature is almost always too high for the hydrate structure to be viable.

       The thickness of this stability zone may be anywhere from 0 to 2 kilometers, and can
expand or contract in response to changing temperature and pressure conditions.  If the zone
expands (as a result of increased pressure and/or decreased temperature),  it can incorporate
more methane from the surrounding sediment. Conversely, if the zone contracts (as a result
                                                                            Page 3-4

-------
of decreased pressure and  or increased temperature), large amounts of  methane can be
liberated from hydrates into the sediment and can migrate into the atmosphere.

3.1.2 Current Emissions

      Oceanic and onshore continental reserves are believed to be stable at present, which
means that they are not currently emitting methane.  However, offshore  continental shelf
reserves  are currently unstable, and  may emit  3  to 5 Tg of methane annually  to the
atmosphere (Kvenvolden 1991e; IPCC 1992). These emissions result from climate changes
that occurred within the last 18,000 years, the time since the last glaciation. Since that time,
sea level has risen about 100 to 125  meters, inundating large areas of permafrost which
contain methane hydrates.  Inundation has increased the pressure in this region by about  9
atmospheres, which would be expected to increase the stability of the hydrates.  However,
over the same period, the temperature at the sediment surface has increased by 15C, which
is more  than  enough to  offset the  increase  in pressure  and destabilize  the  hydrates
(Kvenvolden 1991e).  Due to the slow rates of downward thermal diffusion in sediments, this
temperature change is still in the process of penetrating  downward, degrading the permafrost
and associated gas hydrates. The emission estimate of 3 to 5 Tg CH4/yr was determined by
assuming that the difference between the calculated and the known amount  of methane in
sub-sea permafrost hydrates (160,000 - 43,000 = 117,000 Tg) has been uniformly released
over the last 18,000  years (Kvenvolden 1991e).

      These estimates assume that the methane  being liberated from the  gas hydrate form
is released into the atmosphere.  It is possible, however, that some or all  of this gas is not
actually  emitted  to  the atmosphere.    Instead  it is  oxidized  or absorbed  within the
sediment.24 If the methane is not trapped in sediment or oxidized in the water column,  it
could escape  to the atmosphere and contribute to the atmospheric burden of methane
(Cicerone and  Oremland  1988; Kvenvolden 1991e).

      Estimates of methane release from offshore, continental  shelf gas  hydrates may be
substantiated by measurements of the methane concentration of 17 bottom-sediment samples
from the Alaskan Beaufort shelf (Harrison Bay). Concentrations ranging from 4,000 - 20,000
nl CH4/I water were observed (Kvenvolden et al. 1991 a).  The equilibrium concentration of
CH4 relative to the atmosphere  is only 70 nl/l.  Other researchers have measured bottom-
water methane concentrations as high as 1,100 nl/l H20 on the Canadian  Beaufort shelf, in
an area of known hydrate occurrence. The unusually high methane concentrations observed
in the Beaufort  shelf suggest  a source of methane within sediment of the nearshore
continental shelf. The source could  be gas hydrates (Kvenvolden et al.  1991 a).

      Further evidence that hydrates may currently be releasing methane to the atmosphere
may be provided by Clarke et al. (1986), who suggest that some 150  methane  plumes
observed by NOAA satellites near Bennett Island in the Soviet far Arctic are produced by
hydrates. There are other plausible explanations for these plumes. It has been suggested, for
example, that the plumes may be water vapor-not methane (Kvenvolden, pers.  comm. 1992).
Further research is necessary to confirm Clarke's  hypothesis.
   24 Oxidation is the chemical conversion of methane and oxygen to carbon dioxide and water. Absorption
means that the methane is dissolved in water or trapped in soil and can be oxidized or emitted at a latter date.
                                                                            Page 3-5

-------
3.1.3 Future Emissions

      Due to their proximity to the earth's surface 2000 m), gas hydrates will eventually
be affected by global warming, and methane emissions from this source are likely to increase
if temperature significantly rises.  While pressure on hydrates is also expected to change as
a result of sea level rise and the melting of polar ice caps, temperature changes are likely to
be far more significant than changes in pressure in determining future emissions during global
warming.

      The Intergovernmental  Panel on Climate Change has  predicted that  an  effective
doubling of atmospheric CO2 between 1990 and 2025 to 2050 will cause the global mean
temperature to increase 1.9C to 5.2C (Mitchell et al. 1990).  The temperature increase in
the polar regions, where all onshore and offshore continental shelf hydrates are believed to
occur, is expected to be double the global mean, or about 3C to 10C (IPCC 1990).

      Using a variety of assumptions about the magnitude of temperature rise, gas hydrate
reserves, and the thermal  properties  of oceans and sediment, several researchers  have
developed scenarios for future emissions of methane from this source due to climate change.
The scenarios have been divided into continental and oceanic scenarios, and their key features
are summarized in Exhibits 3-4 and 3-5.  The scenarios predict the expected annual rate of
methane emissions once global  warming has reached the hydrate zone. Most of the  scenarios
also provide a rough estimate  of the time lag expected before warming would  reach the
hydrate  zone.

      Continental Hydrate Scenarios

      The scenarios for the continental hydrates, developed by four different researchers, are
presented below.

      Kvenvolden's Scenario:  150 Tg CH4/yr

       Kvenvolden (1991e) reports that the most likely additional future source of methane
emissions from  hydrates  is the region  of  subsea  permafrost, which may  be  currently
degassing.  The  accelerated temperature rise that would result from the expected climate
change could penetrate to the  level of subsea permafrost and increase the current rate of
emissions from these hydrates by an order of magnitude (from 4-5 Tg CH4/yr to  40-50 Tg
CH4/yr).  This increase in the emission rate would be expected  to take place sometime after
the twenty first century.

      Similarly,   this  scenario predicts that onshore gas  hydrates  will eventually be
destabilized  by climate change.  The rate of emissions from this source is  predicted  to be
twice that of subsea permafrost hydrates, or about 100 Tg CH4/yr. The time lag before this
emission rate is achieved is greater than the lag for subsea permafrost emissions and ranges
from hundreds to thousands of years (see Exhibit 3-5).

      MacDonald's Scenario:  50 Tg CH4 /yr

      Based on the analysis of recent temperature changes in the Arctic by Lachenbruch and
Marshall (1986), MacDonald (1990) assumes that Arctic surface temperature has  risen 2C
                                                                             Page 3-6

-------
since 1880 and that temperature will increase another 2C by 2080. By the year 2090, this
temperature disturbance will have reached the more shallow sections of the hydrate stability
zone and methane release will begin.  McDonald's forecast predicts that the entire continental
hydrate zone will be destabilized at an average rate of 30 cm/yr or about 0.01 % of the total
reserve per year. The total continental reserves are estimated at 5.3x105 Tg CH4/ thus the
annual emissions could be 50 Tg CH4/yr (see Exhibit 3-5).

       Bell's Scenario: 300 Tg CH4 /yr

       Bell (1982) has developed a scenario in which a doubling of atmospheric CO2 causes
the air temperature in the high latitudes to increase by 10C. Consequently, the permafrost
between the -5C and -15C isotherms of annual mean air temperature  becomes  unstable.
As the permafrost is destabilized, temperature rise is transmitted to the  hydrates, which exist
within and below the permafrost zone.   Destabilization and degassing  from the hydrates
begins within a few hundred years of the initial air temperature rise.  The affected area
comprises about half of the total area of northern permafrost, and it also contains about half
of the total continental hydrate reserves (Total continental reserves = 2.7x106 Tg CH4). Bell
estimates that all of the methane in this region will be released uniformly over 4000 years (see
Exhibit 3-5).

       Nisbet's Scenario:  >100Tg CH4/yr

       This scenario is based on the assumptions that CO2 concentration will quadruple within
a century and the mean annual air temperature in the Arctic Islands will increase by 19C,
from -14 to 5C (Nisbet 1989).  Nisbet also makes the extreme assumption that hydrates
are very close to the surface in the Arctic Islands (as close as 50 m). These assumptions lead
to a very high release rate per unit area. However, in this scenario, Nisbet applies this release
rate to only a small part of the total continental hydrate area--for a total emission estimate of
at least 100 Tg/yr.  It  is not clear whether the other hydrate areas are  not destabilized, or if
the scenario is simply not intended to give a global picture.  If, for example, the areal release
rate is applied to the affected area used in Bell's scenario, the annual emissions  from this
scenario would be about 10,000 Tg CH4/yr. In a related paper, Nisbet contends that hydrates
do not exist in all areas of hydrate stability, but only where the surface rocks are sedimentary
-- suggesting that the  total  affected area would  be less than the  area used  by other
researchers (Nisbet and Ingham, submitted). Finally, Nisbet suggests that  global warming will
be transmitted to the hydrate zone in thermal pulses, rather than as a steady process. These
thermal pulses could result in  (1) occasional bursts of methane, as a result of the rupture of
hydrate-trapped gas pools, and (2) slow  methane seepage  to  the  surface  as hydrates
decompose (Nisbet and  Ingham, submitted) (see Exhibit 3-5).
       Oceanic Hydrate Scenarios

       Scenarios of methane emissions from oceanic gas hydrates have been devised by three
different researchers, and are presented below.
                                                                             Page 3-7

-------
      Kvenvolden's Scenario: > 150 Tg CH4/yr

      To place an upper limit on methane emissions from oceanic hydrates, Kvenvolden
makes the rough estimation that  a rise in global temperature will destabilize oceanic reserves
after thousands of years. This will result in methane emissions that could exceed his estimate
of emissions from the  continental  shelf (previously estimated to be  150 Tg CH4/yr).
Kvenvolden  does not provide  a  more specific  estimate  of  oceanic  hydrate  emissions
(Kvenvolden 1991e) (see Exhibit 3-4).
      Bell's Scenario: 160 Tg CH4 /yr
      Bell (1982) assumes that, as a result of global warming, the surface water temperature
of the Norwegian Sea, which feeds the Arctic Ocean, will rise about 3.5C (reflecting a
similar change in air temperature).  When this water enters the Arctic, it descends to the
"intermediate" depth of 250-350 m. The  predicted increase in precipitation in the Arctic
region will increase the extent of mixing with the water entering from the Norwegian Sea.
This warmer, "intermediate" water will extend about halfway around the Arctic Basin.  Thus,
the temperature along half the length of the 300 m depth contour in the Arctic Ocean will
increase from -0.5 to 3C. Given this temperature change, the hydrates extending from the
ocean sediment interface to a depth of 40 m below the sea floor will be destabilized where
ocean depths are between 280  and 370 m.  Bell estimates that the methane in this 40 m
zone, which represents about 1 % of the oceanic methane hydrate reserves, will be entirely
and uniformly released over a 100 year span. To estimate global annual emissions from this
release. Bell assumes that a global oceanic methane reserve of 13  million Tg is  uniformly
distributed in the top 250 m of ocean sediments, at depths between 200 and 1,000 m (see
Exhibit 3-4).

      Revefle's Scenario: 640 Tg CH4/yr

      In this  scenario, Revelle (1983) assumes that  the  mean annual air temperature
increases 3C globally, and that  ocean surface  temperatures rise  by the  same amount.
Advection and eddy currents will carry the heat downward, and bottom water temperatures
will eventually rise 1  to 4C.  As a result, the minimum  depth of hydrate stability will
increase about 100 m at all latitudes, destabilizing the top 100 meters of the hydrate layer
throughout the world.  Revelle assumes that (1) hydrates are evenly distributed within the
hydrate zone, and (2) that 20% of the released methane will be absorbed by the  water in the
pores of the ocean-bottom sediment before it can reach the atmosphere.  He calculates that
after approximately 100 years, oceanic methane hydrates will begin  emitting methane at an
annual rate of 640 Tg. Due to its magnitude, this feedback alone could result in as much as
2C of additional global warming (Revelle 1983) (see Exhibit 3-4).

      Composite Scenarios

      It is  difficult  to compare  these scenarios  because they are  based on different
assumptions of the expected temperature  rise and of  the  size  of  hydrate  reserves.  To
normalize the assumptions about temperature rise and hydrate reserves across all scenarios,
an "emission factor" is calculated here for each scenario:
                                                                             Page 3-8

-------
Exhibit 3-4
Oceanic Hydrate Scenarios

Air Temperature Rise (C)
Methane Reserves (Tg)
% of Area/Reserves
Destabilized
Destabilized Area (m2)
Time Until Destab. Begins (yrs)
Time to Fully Destabilize (yrs)
Efficiency of Escape to
Atmosphere
Emission Factor (yr'1C'1)
Annual Emissions (Tg/yr)
Kvenvolden
3*
1x107


1000's
1000's
100%
5.0x10'6
>150
Revelle
3
1.8x107
13%
1.9x1012
100

80%
1.2x10'5
640
Bell ||Composite
3-4
1.3x107
1%
1.2X1011
100's
100
100%
3.4x1 0'6
160
3
1x107




100%
6.7x10'6
200
This value was not provided by the researcher and was assumed to be
equal to the composite scenario value.
      The emission factors for the above scenarios are presented in Exhibits 3-4 and 3-5,
along with average emission factors for all oceanic and continental scenarios.  There appears
to be general agreement regarding these factors, within an order of magnitude. The existing
difference in emission factors primarily represents different  conceptions of the thermal
conductivity of sediment and of  the percentage of total hydrates that will be affected by
global warming.

      The average emission factors can  be applied  to the  current "best  estimates"  of
temperature rise and hydrate reserves to arrive at composite, or best-estimate, scenarios.  In
the oceanic composite scenario, the temperature rise is assumed to be equal to the midpoint
of the IPCC global prediction (3C), since oceanic hydrates  are globally distributed. For the
continental composite scenario, the temperature rise is assumed to be equal to the average
IPCC prediction for the polar regions (6C), since  continental hydrates exist only at high
latitudes.  The  total reserves are assumed to be equal to the  latest estimates provided by
Kvenvolden  (1x107 Tg for oceanic, 5x105 Tg for continental).

      The composite  scenarios  predict annual emissions  of  about 200  Tg CH4/yr from
oceanic hydrates and  about 100 Tg CH4/yr from continental hydrates with a  range  of
uncertainty of at least one order  of magnitude. These composite  scenarios are intended  to
provide a rough, average estimate of the magnitude of hydrate  releases. However, some  of
the emission factors in the researchers'  scenarios are intentionally extreme.  On the other
hand, the temperature rise and reserve estimates assumed in the composites are not intended
to be extreme, but do encompass a large degree  of uncertainty.  The annual emissions
                                                                             Page 3-9

-------
predicted by the composites could be orders of magnitude larger or smaller depending on the
temperature rise and reserve assumptions used.

      It should also be noted that the composite scenarios assume that the efficiency of
escape to the atmosphere is 100%. This assumption is made because only one of the seven
hydrate release scenarios mentions that the rate of absorption  or oxidation is  not 100%.
However, there is reason to believe that efficiency could be less than 100% (see section 3.3).
Therefore, this is somewhat of an  extreme assumption because efficiency can not be more
than 100% but could be much less than 100%.
Exhibit 3-5
Continental Hydrate Scenarios

Air Temperature Rise (C)
Methane Reserves (Tg)
% of Area/Reserves
Destabilized
Destabilized Area (m2)
Time Until Destab. Begins (yrs)
Time to Fully Destabilize (yrs)
Efficiency of Escape to
Atmosphere
Emission Factor
(yr-1C1)
Annual Emissions (Tg/yr)
Kvenvolden
(offshore) (onshore)
6* !6*
I
1.5x105 ''3.5x1 05
i
i
i
i
i
i
i
i
1
100's '1000's
1000's '1000's
I
100% |100%
1
5x10'5
50 ;100
MacDonald
2
5.3x1 05
100%

100
10,000
100%
4.7x1 0'5
50
Bell
10
2.7x106
50%
6.5x1012
100's
4000
100%
1.3x10-5
300
Nisbet
19
5x1 O5*

1.0x1011
100-200

100%
1.1x10'5
220
Composite
6
5x1 05




100%
3.3x1 0-5
100
*These values were not provided by the researchers and were assumed to be equal to the
composite scenario values.
3.2    Permafrost

       Permafrost is ground that remains at or below 0C throughout the year for at least two
consecutive years, and usually contains ice held  within soil pores.  Methane can occur in
permafrost in ice, peat, loess, and any material that is perennially frozen. It is believed that
some, if not all, permafrost contains trapped methane. While little research has  been done
on the subject of fossil emissions of methane from permafrost, this is a significant, potential
source of atmospheric methane in the future.
                                                                           Page 3-10

-------
      Permafrost methane is not the same as methane hydrates. Permafrost methane is
methane with the standard molecular structure that is trapped within permafrost.  Methane
present where the hydrate stability zone overlaps with the permafrost zone (see exhibit 3-3),
exists in the form of methane hydrate, and is not considered permafrost methane.

3.2.1 Background

      To draw reasonable conclusions about methane  emissions from  permafrost, it is
important to understand the following characteristics of permafrost:

            The amount of  methane in permafrost
            The rate at which permafrost is melting now and can be expected to melt in the
             future
            The total area of permafrost subject to melting
            How much of the methane released from permafrost can be expected to reach
             the atmosphere.

These characteristics are discussed below.

      Methane Concentration and Total Reserves

      Methane concentration is the weight  of  methane per unit weight  of  permafrost.
Kvenvolden (1991d) has measured the methane content of shallow permafrost cores from
three sites on the campus of the University of Alaska, Fairbanks. Methane concentration was
highly variable, ranging from 0.00032 to 22 mg CH4 per kg of sample. Core samples taken
from a permafrost tunnel excavated by the U.S.  Army Corps of Engineers at a site  16 km
north of Fairbanks, Alaska, contained 0.0008 to 6.6 mg CH4 per kg of sample (Kvenvolden
1992). The number of samples analyzed in both of these experiments was fairly small, and
the average concentration is still uncertain.

      A preliminary estimate  of partial reserves of methane in permafrost can be extrapolated
from the known volume of ice in permafrost. There are approximately 250,000 km3 of ice
worldwide,  of which 80% is ground ice within  permafrost (Michael  Smith, pers.  comm.
1992).  The  total mass  of  permafrost  ice is  then 2.5x1017  kg.   Applying the  above
concentration estimates to this estimate of the volume of permafrost ice results in a range for
partial methane reserves of 0.08 to 5,500 Tg CH4.  Additional methane reserves could exist
in the non-ice portions of permafrost, the total mass of which is not known at this time.

      Destabilization Rate

      The destabilization rate is the rate at which permafrost melts, usually expressed as
downward distance from the  top edge of the permafrost per unit time.  Permafrost is stable
where the mean  annual surface temperature (MAST) is less than 0C. Permafrost stability
is also related to vegetation, seasonal snow cover, geological setting,  and topography.
Osterkamp postulates that when MAST rises above 0C, Alaskan permafrost typically melts
at a rate of 10-20 cm/yr, while the time to achieve a new equilibrium thickness can be less
than one hundred years or more than ten thousand  years (Osterkamp, submitted).
                                                                          Page 3-11

-------
      Destabilization Area

      The destabilization area is the areal extent of permafrost subject to destabilization, and
is a subset of the total area of permafrost.  Permafrost exists beneath approximately one
quarter of the land area of the earth, or 3.6 x 1013 m2 (Michael Smith, pers. comm. 1992).

      Efficiency

      Efficiency refers to the percentage of methane released from permafrost that actually
reaches the atmosphere, as opposed to being  oxidized, in the active soil  layer.  The  most
critical factor affecting permafrost methane  emissions will likely be  the  oxidation rate of
permafrost methane as it makes its way through the active soil layer.

      While no research has been published on the  efficiency of methane escaping  from
melting permafrost, studies  of methane released from other sources (e.g., landfills, wetlands)
suggests that oxidation rates can be very high (Whalen et al. 1990). By measuring methane
concentrations in pore water and atmospheric emissions from northern wetlands, it has been
demonstrated that the vast majority of methane produced in relatively well-drained soils is
consumed before it reaches the atmosphere (Whalen and Reeburgh  1992).  On the  other
hand, participants at the workshop held at Oak Ridge (see section 2.5.3) hypothesized that
permafrost methane  "should readily escape [to the atmosphere] if permafrost melts"  (Post
1990).

      The largest single determining factor of oxidation rate is soil moisture.  However,
conflicting hypothesis exist  for how oxidation will be affected by soil moisture.  For example,
some researchers contend that if  the soil is  completely inundated (the water table is at or
above the soil surface) then  the released methane is likely to pass through this layer and reach
the atmosphere (high efficiency). Others contend that "if the active soil layer becomes wetter
then the rate of methane transport through it will be slowed and the probability of oxidation
could increase, assuming the soil does not become so wet that it goes anaerobic" (Mulkey,
pers. comm. 1993).

      In a drier scenario it has been suggested that if a well-drained layer exists, even as little
as a few centimeters, then  conditions are ripe for  oxidation, and some or all of the released
methane could be oxidized in this zone (low efficiency) (Bartlett, pers. comm. 1993).  It has
also been suggested that "if the  soil becomes drier then the rate of transport should be
increased and emissions could become more likely, particularly if the active layer becomes so
dry that bacterial growth is inhibited" {Mulkey, pers. comm. 1993).

3.2.2 Current Emissions

      The factors described above can be combined to quantitatively calculate current and
future emissions as follows:

      emissions = (concentration) x (destabilized area) x (destabilization rate) x (efficiency)

Unfortunately, there is insufficient information at this time to apply this formula to current or
future emissions.  Information  does exist, however, to allow the formation of some general
hypotheses about  permafrost methane emissions.
                                                                             Page 3-12

-------
      Evidence exists to suggest that temperature in the  polar regions has already risen
considerably since pre-industrial times, and that permafrost is melting and releasing methane.
A detailed analysis of sub-surface temperature records from wells drilled in Alaska indicates
that the  permafrost surface temperature has risen by 2  to 4C over the past century
(Lachenbruch and Marshall 1986). Furthermore, some discontinuous areas of permafrost in
Alaska are known to be melting at this time (Osterkamp 1983). No estimates are available
as to what quantity of permafrost is melting, or how much methane is being liberated in this
process.

3.2.3 Future Emissions

      Permafrost will be destabilized  by global warming in the future as a result of expected
increases in temperature. While the relationship between permafrost temperature and surface
temperature is not well defined, it has  been hypothesized that if global temperatures increase
by only 1 C, then permafrost will melt entirely in 8% to 20% of the current permafrost area,
and  it will recede by an average  of 0.5 m  in the remaining area (Osterkamp,  pers. comm.
1992). With a temperature increase of 2C, 20% to 40%  of the permafrost area will be lost,
and  the top 1 m of permafrost will melt in the  remaining areas (Osterkamp,  pers. comm.
1992).  However, there may be some counteracting effects.  For example, an increase in
snow cover  could reduce the above predictions by  as much as 50% (IPCC 1990).

      Methane will be released as permafrost melts due to  climate change.  The amount of
methane that will be liberated, and how much of this will actually reach the atmosphere, is
not  known.  Currently, no scenarios exist in the literature for future  emissions from
permafrost.  Preliminary results from Khalil's ongoing study of potential permafrost emissions
show that as much as 60 Tg CH4 per year could be released to the atmosphere by the end
of the next century (M.A.K. Khalil, pers. comm.  1992).
3.3    Uncertainty / Further Research

       Knowledge of natural methane emissions from fossil sources is limited. Therefore, a
large degree of uncertainty must be attached to future emission scenarios.  The uncertainty
could be partially resolved with additional research. Uncertainty exists in the following areas:

       Methane Reserve Estimates

       The total quantity of methane currently stored in the  form of gas hydrates or
permafrost methane is not well  known.

       Gas Hydrates

       Estimates of the exact quantity of methane hydrate reserves vary by over three orders
of magnitude for oceanic hydrates and by two orders of magnitude for continental hydrates.
While recent, independent estimates agree on the values provided by Kvenvolden (1991e),
these estimates are still highly uncertain.  An extensive program of  sediment sampling on
continental  slopes and  in  permafrost regions worldwide, to determine  hydrate  depth,
thickness, distribution, and methane concentration could considerably reduce this uncertainty.
                                                                            Page 3-13

-------
      Permafrost

      Very few samples of permafrost have been tested for methane concentration, and all
of these have been from a very small geographic area around Fairbanks, Alaska. A much
more intensive and geographically diverse permafrost sampling program is needed. Also, a
methodology must be devised for estimating the global volume of permafrost, and not just
permafrost ice.
      Hydrate Distribution

      In general, the scenarios for  gas hydrates presented  above assume that hydrate
reserves are uniformly distributed within the hydrate zone. However, this assumption is not
well substantiated. Kvenvolden (1988) points to exploratory drilling results from Prudhoe Bay
and the coast of Guatemala as evidence that "hydrates are present in discrete layers and
occupy  only a small part of the potential field of gas hydrate stability."  If hydrates are not
prevalent throughout the projected zones of destabilization, the scenarios may overstate the
potential from this source.  However,  if the  hydrates are discontinuous  but fairly evenly
distributed  by depth  and latitude, the scenarios may  be  accurate. A program of sediment
sampling will help resolve this issue.

      Thermal Conductivity

      Thermal conductivity (the ability of the water/soil column to transmit  changes in
surface  temperature) will determine the rate at which  surface warming is conducted to, and
through, the hydrate and permafrost  zones. Thus, thermal properties of water and soil will
affect not only the time lag before destabilization begins, but also how fast destabilization will
take place--the annual emission rate. However, thermal properties should affect only the time
frames  involved  before a  new equilibrium stability is reached--not the total magnitude of
destabilization.

      Gas Hydrates

      In the oceanic hydrate case, thermal conductivity is determined not only by the thermal
properties of seawater and sub-ocean  sediment,  but also by  ocean mixing, which can be
highly variable and difficult to determine.  For example, in Bell's oceanic scenario,  global
atmospheric temperature rise rapidly affects (over a span of about 100 years) the bottom
temperature of the Arctic Ocean at a depth of 300 m. This rapid  translation of surface
warming is based on assumptions of water circulation patterns.  According to  Kvenvolden
(1988), however, this scenario overestimates the degree of mixing between the Arctic Ocean
and the Canada Basin and, thus, the time delay before the Arctic Ocean bottom temperature
would  rise.  Ocean  circulation  patterns must be more  clearly understood  to  resolve this
uncertainty.

      Thermal conductivity is better understood in the continental hydrate case than in the
oceanic case, but it is still difficult to  model.  Conductivity  depends  on soil  porosity,
conductivity of soil  grains, and whether the  soil is frozen or thawed.  Determining these
conditions can be difficult.  For instance, even below 0C, sediment may not be frozen, due
to surface effects that are  not well understood at this time (MacDonald 1990). Furthermore,
                                                                             Page 3-14

-------
different amounts of energy are required to melt ice and hydrate. Movement of the ice/hydrate
boundary, therefore, complicates the calculation of thermal conductivity.

      Due to the wide range  of  possible conditions  for thermal  conductivity,  some
researchers have concluded that huge annual methane hydrate emissions are possible within
a century, and others have concluded that annual emissions will be relatively small and will
only begin to take place after thousands of years. It is generally agreed, however, that the
methane hydrate reserves with  the shortest lag time before destabilization begins are the
offshore continental shelf reserves (Kvenvolden 1991e; Lashof, pers. comm.).

      Permafrost

      The time lag before temperature rise reaches permafrost methane is both much shorter
and more certain than for gas hydrates because permafrost exists much closer to the surface,
and there is less room for variability.  However, understanding  of the rate  of permafrost
destabilization is still limited.   Further research on the  current distribution  of permafrost
surface temperatures and how these relate to air temperatures is necessary to make a realistic
assessment  of the effects of a given increase in  air temperature on permafrost stability
(Osterkamp, pers. comm. 1992).

      Methane Oxidation / Absorption

      Another potentially mitigating factor for both the permafrost and gas-hydrate scenarios
is the amount of escaping methane that will either be absorbed  by the water in the pores of
the surrounding sediment, or oxidized in the water column, or the active soil layer, before it
reaches the atmosphere. Absorption means that the methane becomes dissolved in water or
trapped  in soil and can eventually come out of solution and reach the atmosphere. Revelle
assumes that 20% of the methane released from oceanic hydrates is trapped in sediment
before it reaches the ocean.

      Methane is oxidized when it reacts with oxygen in soil or water and is converted to
carbon dioxide and water. For terrestrial hydrates, the rate of oxidation depends primarily on
soil moisture,  precipitation, and soil porosity.  No estimates have  been  made for the
percentage of methane released from continental hydrates that will be oxidized. Oxidation of
oceanic  hydrate methane depends on the oxygen content of the water.  Little information is
known about the rate of oxidation of methane from hydrates in the water column, but Revelle
(1983) contends that all methane that escapes from the sea  floor sediment "should rise
rapidly to the sea surface before it can be oxidized in the water". Other hydrate modelers
make the assumption, perhaps arbitrarily, that released methane is  neither trapped nor
oxidized in sediment or the water column, but escapes to the atmosphere.

      Studies of methane released from non-hydrate sources show that oxidation in both soil
and water columns is  very significant.  Research has shown that much  of the methane
generated in landfills and wetlands can be oxidized in the soil column {Whalen et al.  1990;
Whalen  and Reeburgh  1992).   Similarly, geochemical studies  have shown that dissolved
methane, emanating for example from hydrothermal vents, is rapidly oxidized in the water
column by bacteria (Mulkey, pers. comm. 1993). Whether methane released from hydrates
or permafrost will experience similar oxidation rates to these other sources is not known and
therefore needs to be researched. While it may be possible to design experiments that would
                                                                           Page 3-15

-------
help resolve this uncertainty, it is also possible that methane will be released from hydrates
in the future on such a large scale that there is no reliable way of predicting the oxidation
processes that would be involved.
                                                                                Page 3-16

-------
                                    CHAPTER 4

                          SUMMARY AND CONCLUSIONS
      Current emissions from natural wetlands have been reasonably well quantified, and
experts  have agreed on annual  emissions of approximately 109 Tg of methane, with an
estimated range of 70 to 170 Tg of methane per year. Emissions from fossil sources are less
well known, but probably contribute only marginally to atmospheric methane at the present
time.

      It is difficult to predict future methane emissions from natural sources or to assess the
potential positive  feedback contribution of methane emissions from natural sources to the
climate system.  However, experts in this field have attempted to  estimate the amount of
methane that might be expected from these sources and over what time periods these
changes may occur, assuming that current trends in energy use and agricultural activities
continue. The IPCC predictions of global warming are used as standard assumptions for these
"best guesses" (see Exhibit 4-1).

      In order to  predict future  methane emissions  from  wetlands,  researchers  have
attempted to quantify the relationship between wetland emissions and the environmental
variables that will be affected by climate change.  Relationships have been hypothesized
between emissions and temperature, precipitation as it affects inundation, and plant species
and  growth  rates.   Rough methods of calculating  future  emissions  based  on  these
relationships have been developed for northern wetlands, but are difficult to discern from the
limited and highly variable data available from tropical wetlands.

      Given the relationships between methane emissions from wetlands and environmental
variables, researchers have concluded that it is unlikely that emissions will decrease as a result
of climate change over the next  century,  instead,  wetland emissions will probably increase.
Quantitative predictions of potential emissions increases for northern wetlands, based on three
scenarios of climate change, range widely from 5 Tg/year to 290 Tg/year.

      The rise in methane emissions from wetlands would take place roughly concurrently
with climate  change. Fast and slow mechanisms have been identified by which climate
change will alter methane emissions. For example, the emission rate of a particular wetlands
ecosystem will be affected immediately by climate change, while the area! extent and plant
community will be  affected gradually, over decades.  According  to these mechanisms,
methane emissions from wetlands will  likely reach the predicted increased levels within the
next century.

      By all accounts, fossil sources could release tremendous quantities of methane, some
of which could reach  the atmosphere.  These  emissions could amount to hundreds of
teragrams annually from hydrates,  and 60 Tg/yr from permafrost.  However, the timing of
these fossil emissions is highly uncertain.  While permafrost emissions could occur within a
century, major hydrate emissions  are  not likely  to take place for at least one to several
hundred years, and  may not occur at all within the next few thousand years.  Importantly,
even though these emissions may not  occur for many years, they would still be a result of
current day activities and their related emissions of greenhouse gases.
                                                                             Page 4-1

-------
Exhibit 4-1
Potential Increases in Methane Emissions from Selected Natural Systems due to
Climate Change
System
Tropical Wetlands
Temperate Wetlands
Northern Wetlands
Hydrates (Continental)
Hydrates (Oceanic)
Permafrost
Total
Current Emissions
(Tg CH4/yr)
estimate
66
!5
38
5
0
0
114
range
34-118
n/a
29-52
0-5
0
0
70- 175
Range of Predicted Increases
(Tg CH4/yr)
within 1 00 yrs
L <>1
O1
5-290
O2
O2
0-602
5 -370
after 100 yrs
O1
O1
5 -290
50 - 3002
150-6402
0-602
200- 1300
1 In the absence of future predictions for these systems, they are assumed here to
remain constant.
2 The predicted increases for hydrates and permafrost are relatively more uncertain
than the predicted increases for wetlands.
      To summarize, while future emissions predictions are highly uncertain, especially for
fossil sources, available information indicates that methane emissions from all natural sources
could increase by 5 to 370 Tg/year within the next century. This increase would be primarily
from wetlands and  permafrost, which respond to climate change more  quickly than gas
hydrates.  The methane from such an increase is equivalent to about 100 to 7,000 million
metric tons of CO2 per year, or an additional 0.2% to 17% of carbon dioxide beyond current
predictions for the year 2100. In the centuries to follow the next one, methane emissions
from natural sources could rise further as gas hydrates are slowly destabilized.

      With additional research,  scientists will be able to more confidently predict  future
natural source emissions. For wetlands, additional multi-year flux studies and additional work
in more geographical areas will reduce the uncertainty surrounding current emission estimates,
and improve understanding of the relationships between methane and environmental variables.
More complex wetland models need to be developed as well, incorporating as many climate
variables as possible. Future predictions can only be accurate if they are based on correct
assumptions  of  climate change.  Therefore,  further  efforts to increase the accuracy and
resolution  of general  circulation models will  also  improve  natural  methane emissions
projections.
                                                                              Page 4-2

-------
       Scientists have only recently begun to investigate the potential of gas hydrates and
permafrost methane to contribute to global warming.   As this field  is more thoroughly
explored, a more accurate picture will develop. This process will require improved estimates
of the magnitude and distribution of hydrate and permafrost methane reserves, as well as a
greater understanding of the thermal properties of permafrost and of the ocean waters and
soil layers in the vicinity of hydrates. Even if these issues are resolved, it may not be possible
to determine what percentage of released methane will actually reach the atmosphere until
large, fossil-source emissions are in progress.
                                                                              Page 4-3

-------
                                           REFERENCES
Aselmann, I. and P.J. Crutzen, Global distribution  of natural freshwater wetlands and rice paddies, their net
primary productivity, seasonality and possible methane emissions, J. Atmos. Chem.. 8, 307-358, 1989.

Baker-Blocker, A., T.M. Donahue, and K.H. Mancy, Methane flux from wetlands areas, Tellus. 29.245-250,1977.

Barber, T.R., R.A. Burke, and W.M. Sackett, Diffusive flux of methane from warm wetlands. Global Bioaeochem.
Cycles. 2.411-425. 1988.

Bartlett, K.B.,  P.M. Grill,, J.A. Bonassi, J.E.  Richey, and R.C. Harriss, Methane flux from  the  Amazon River
floodplain: Emissions  during rising water, J. Geophvs. Res..  95, 16,773-16,788, 1990.

Bartlett, K.B., P.M. Grill, D.I. Sebacher, R.C. Harriss, J.O. Wilson, and J.M. Melack, Methane flux from the central
Amazonian floodplain, J. Geophvs. Res.. 93. 1571-1582, 1988.

Bartlett, K.B., P.M. Grill,  R.L. Sass, R.C. Harriss, and N.B. Dise, Methane emissions from tundra environments in
the Yukon-Kuskokwim Delta, Alaska, J. Geophvs. Res.. 97, 16645-16660, 1992.

Bartlett, K.B., R.C. Harriss, and D.I. Sebacher, Methane flux from coastal salt marshes, J. Geophvs. Res.. 90.
5710-5720, 1985.

Bartlett, K.B.,  D.S. Bartlett, R.C.  Harriss, and  D.I.  Sebacher, Methane emissions along a  salt  marsh salinity
gradient,  Bioaeochem.. 4, 183-202,  1987.

Bartlett, D.S., K.B. Bartlett, J.M. Hartman, R.C. Harriss,  D.I. Sebacher, R. Pelletier-Travis, D.D. Dow, and D.P.
Brannon, Methane emissions from the Florida Everglades:  Patterns of variability in a regional wetland ecosystem,
Global  Biooeochem. Cycles. 3, 363-374, 1989.

Bell, P.R., Methane Hydrate and the Carbon Dioxide Question, in Carbon Dioxide Review. 1982 , edited by W.C.
Clark, pp. 401-406, Oxford University Press, New York,  1983.

Bell, P.R., Methane hydrate and the carbon dioxide question, in: Carbon Dioxide Review. 1982. edited by W.C.
Clarke, Oxford Univ. Press, N.Y.,  pp. 401-406, 1983.

Billings, W.D.,  J.O. Luken, D.A. Mortensen, and K.M. Peterson, Arctic tundra: A source or sink for atmospheric
carbon dioxide in a changing environment?, Oecoloaia (Berl). 53. 7-11, 1982.

Blake, D.R. and F.S. Rowland, Continuing worldwide increase in tropospheric methane, 1978 to 1987. Science.
239. 1129-1131, 1988.

Born, M., H. Dorr, and I. Levin, Methane consumption in aerated soils of the temperate zone, Tellus. 42B. 2-8,
1990.

Burke,  R.A.,  T.R. Barber, and W.M. Sackett, Methane flux and stable hydrogen and carbon isotope composition
of sedimentary  methane from the Florida Everglades, Global Bioaeochem. Cycles. 2, 329-340, 1988.

Chanton, J.P.,  C.S. Martens,  C.A. Kelley, P.M. Grill, and W.J. Showers, Methane transport mechanisms and
isotopic fractionation  in emergent macrophytes of an Alaskan tundra lake, J. Geophvs. Res..  97, 16627-16643,
1992.

Chanton, J.P., G.J. Whiting, W.J. Showers, and P.M. Grill. Methane flux from  Peltandra virginica: Stable isotope
tracing and chamber effects. Global Bioaeochem. Cycles. 6, 15-31, 1992.

Chanton, J.P. and J.W.H. Dacey, Effects of vegetation on methane flux, reservoirs and carbon isotopic composi-
tion, in Environmental and Metabolic Controls on Trace Gas Emissions from Plants, edited by T. Sharkey, E.
Holland, and H. Mooney, Academic Press, New York, pp. 65-92, 1991.
                                                                                           Page RF-1

-------
Chappellaz, J.  et al., The Atmospheric CH4 Increase since the Last Glacial Maximum: 1.  Source Estimates,
submitted to Tellus. 1992.

Chappellaz, J. etal., Nature. 345, 127, 1990.

Cicerone, R.J., and R.S. Oremland, Biogeochemical Aspects of Atmospheric Methane, in Global Biogeochemical
Cycles, 2, 299-327, 1988.

Cicerone, R.J. and R.S. Oremland, Biogeochemical aspects of atmospheric methane, Global Biogeochemical Cycles.
2, 299-327, 1988.

Clarke, J.W. et al..  Possible Cause of Plumes from Bennett Island, Soviet Far Arctic, in Am. Assoc.  Petrol. Geol.
Bull., 70, 574, 1986.

Clarke, J.W., P. St. Amand, and M. Matson, Possible cause of plumes  from Bennett Island, Soviet Far Arctic,
Amer. Assoc. Petroleum Geol. Bull.. 70. 574, 1986.

Clymo, R.S. and E.J.F. Reddaway, Productivity of Sphagnum (Bog-moss) and peat accumulation, Hidrobiologia.
12, 181-192, 1971.

Conway, T.J. and L.P. Steele, Carbon dioxide and methane in the Arctic atmosphere, J. Atmos. Chem.. 9,81 -100,
1989.

Craig, H. and C.C. Chou, Methane: The record in polar ice cores,  Geophysical Res. Lett.. 9,  1221-1224, 1982.

Crill, P.M., K.B. Bartlett, R.C. Harriss, E. Gorham, E.S. Verry, D.I. Sebacher, L. Madzar, and W. Sanner, Methane
flux from Minnesota peatlands. Global Biogeochem. Cycles. 2, 371-384, 1988.

Crill, P.M., Seasonal patterns of methane uptake and carbon dioxide release by a temperate woodland soil, Global
Biogeochem. Cycles. 5, 319-334, 1991.

Crill, P.M., K.B. Bartlett, J.O. Wilson, D.I. Sebacher,  R.C. Harriss, J.M. Melack, S. Maclntyre, L. Lesack, and L.
Smith-Morill, Tropospheric methane from an Amazonian floodplain lake, J. Geophvs. Res.. 93,1564-1570,1988.

Crutzen, P.J., Methane's  Sinks and Sources, in Nature,  No. 350, April 1991.

Dacey, J.W.H., Knudsen-transitional flow and gas pressurization in leaves of Nelumbo.  Plant Phvsiol.. 85. 199-
203,1987.

Dacey, J.W.H., Internal winds in water lilies: An adaptation for life in anaerobic sediments. Science. 210. 1017-
1019, 1980.

Dacey, J.W.H. and M.J. Klug, Methane efflux from lake sediments through water lillies, Science. 203,1253-1254,
1979.

DeLaune, R.D., C.J. Smith, and W.H. Patrick, Methane release from Gulf Coast wetlands, Tellus. 35B. 8-15,1983.

Delmas, R.A., J. Servant, J.P. Tathy, B. Cros, and M. Labat,  Sources and sinks of methane and carbon dioxide
exchanges in mountain forest in equatorial Africa,  J. Geophvs. Res..  97. 6169-6179, 1992a.

Delmas, R.A., A. Marenco, J.P. Tathy, B. Cros, and J.G.R. Baudet, Sources  and sinks of methane in the African
savanna: CH4 emissions from biomass burning, J. Geophvs. Res.. 96. 7287-7299, 1991.

Delmas, R.A., J.P. Tathy, and B. Cros, Atmospheric  methane budget  in Africa, J. Atmos. Chem.. 14. 395-409,
1992b.

Devol, A.M., J.E. Richey, W.A. Clark, S.L. King, and  L.A. Martinelli, Methane emissions to the troposphere from
the Amazon floodplain, J. Geophvs.  Res.. 93,  1583-1592, 1988.
                                                                                           Page RF-2

-------
Devol, A.M., J.E. Richey, B.R. Forsberg, and L.A.  Martinelli, Seasonal dynamics in methane emissions from the
Amazon River floodplain to the troposphere, J. Geophvs. Res.. 25, 16,417-16,426, 1990.

Dise, N.B., Methane emission from Minnesota peatlands: Spatial and seasonal variability, Global Bioaeochem.
Cycles, in press (a).

Dise, N.B., Winter fluxes of methane from Minnesota peatlands, Bioqeochem.. in press (b).

Dise, N.B., Methane emission from peatlands in northern Minnesota, Ph. D. dissertation, University of Minnesota,
138 pp.,  1991.

Dobrynin, V. M., et al., Gas Hydrates: A possible energy resource, in Long-Term Energy Resources, vol. 1, edited
by R.F. Meyer and J.C. Olson, pp. 727-729, Pitman, Boston, MA., 1981.

Dooge, J.C.I.,  Hydrologic models and climate change, J. Geophvs. Res.. 97. 2677-2686, 1992.

Edwards,  G., H.H. Neumann, G. den  Hartog, 0. Thurtell,  and G.E.  Kidd, Eddy correlation measurements of
methane  fluxes using a tunable  diode laser at the Kinosheo Lake tower site  during the Northern  Wetlands
Experiment, EOS. 72, 85, 1991.

EPA, Policy options  for stabilizing global  climate:  report to Congress,  D.A.  Lashof, D.A. Tirpak,  eds., US
Environmental  Protection Agency, Office of Policy, Planning and Evaluation, December, 1990.

EPA, International Methane Emissions: report to Congress, US Environmental Protection Agency, Office of Policy,
Planning  and Evaluation, Washington, D.C., (in preparation) 1993.

Fallen, R.D., S. Harrits, R.S. Hanson,  and  T.D. Brock, The role of methane  in internal carbon  cycling  in Lake
Mendota  during summer stratification,  Limnol. Oceanogr., 25. 357-360, 1980.

Fan, S-M., S.C. Wofsy, P.S. Bakwin, D.J. Jacob, S.M. Anderson, P.L. Kebabian, J.B. McManus, C.E.  Kolb, and
D.R. Fitzjarrald, Micrometeorological measurements of CH4 and CO2 exchange between the atmosphere  and the
arctic tundra, J. Geophvs. Res.. 97, 16627-16643, 1992.

Fechner,  E.J. and H.F. Hemond, Methane transport and oxidation in the unsaturated zone of a Sphagnum peatland,
Global Biogeochem. Cycles. 6, 33-44, 1992.

Fung, I.,  J. John, J. Lerner, E. Matthews,  M. Prather,  L.P. Steele, and P.J. Fraser, Three-dimensional model
synthesis of the global methane cycle, J. Geophvs.  Res.. 96. 13,033-13,065, 1991.

Glaser, P.H., G.A. Wheeler, E. Gorham,  and H.E. Wright, Jr., The patterned mires of the Red Lake Peatland,
northern  Minnesota: Vegetation,  water chemistry, and landforms,  J. Ecol., 69.  575-599, 1981.

Goreau, T.J. and W.Z. de Mello,  Effects of deforestation on sources and sinks of atmospheric  carbon  dioxide,
nitrous oxide, and methane  from  Central Amazonian soils and biota during the dry season: A preliminary study,
in: Biogeochemistry of Tropical Rain Forests: Problems for Research, edited by D. Athie, T.E. Lovejoy, and P. de
M. Oyens, Centro do Energia Nuclear na Agricultura and World Wildlife Fund, Piricicaba, S.P., Brazil, pp. 51-66,
1985.

Greco, S., R. Swap, M. Garstang, S. Ulanski, M. Shipham, R.C. Harriss, R. Talbot, M.O. Andreae, and  P.  Artaxo,
Rainfall and surface kinematic conditions over central Amazonia during ABLE 2B, J. Geophvs. Res.. 95. 17,001-
17,014,  1990.

Grosse, W. and J. Mevi-Schutz, A beneficial gas transport system in Nvmphoides peltata. Amer. J. Bot.. 74. 947-
952, 1987.

Guertin,  D.P.,  P.K. Barten, and K.N. Brooks, The peatland hydrologic model: Development and testing, Nordic
Hvdrol.. IfJ, 79-100, 1987.
                                                                                            Page RF-3

-------
Guthrie, P.O., Biological methanogenesis and the CO2 greenhouse effect. J. Geophvs. Res.. 91. 10,847-10,851,
1986.

Hameed, S.  and R. Cess, Impact of a global warming on biospheric sources of methane and its climatic
consequences, Tellus. 35B. 1-7, 1983.

Hamilton, J.D., C.A. Kelley, and J.W.M. Rudd, Methane and carbon dioxide flux from ponds and lakes of the
Hudson Bay Lowlands, EOS. 72. 84, 1991.

Hao, W.M., D. Scharffe, P.J. Crutzen, and E. Sanhueza, Production of N2O, CH4, and CO2 from soils in the tropical
savannah during the dry season, J. Atmos. Chem.. 7., 93-105, 1988.

Harriss, R.C.,  D.I. Sebacher, and P.P. Day, Methane flux in the Great Dismal Swamp, Nature. 297. 673-674,1982

Harriss, R.C., E. Gorham,  D.I. Sebacher, K.B. Bartlett, and P.A. Flebbe, Methane flux from northern peatlands,
Nature. 315.  652-654, 1985.

Harriss, R.C. and D.I. Sebacher, Methane flux in forested freshwater swamps of the southeastern United States,
Geophvs. Res. Lett.. 8, 1002-1004,  1981,

Harriss, R.C., D.I. Sebacher, K.B. Bartlett,  D.S. Bartlett, and P.M. Crill, Sources of atmospheric methane in the
south Florida  environment. Global Biogeochem. Cycles. 2, 231-243, 1988.

Harriss, R.C. and S. Frolking, The sensitivity of methane emissions from northern freshwater wetlands to global
change, in: P. Firth and S. Fisher (ed.s), Global Warming and Freshwater Ecosystems, pp.48-67,  1992.

Harriss, R.C.,  G.W. Sachse, G.F. Hill, L. Wade, K. Bartlett, J. Collins, L.P. Steele, and P. Novell!, Carbon monoxide
and methane in the North American Arctic and Subarctic troposphere, J. Geophvs. Res.. 97,16589-16599,1992.

Harriss, R., K. Bartlett, S. Frolking, and P.  Crill, Methane emissions from northern high latitude wetlands, Proc.
10th Int. Symp. on Env. Biogeochem., 19-23 Aug. 1991, San  Francisco, CA, in press, 1992.

Hill,  P.R., et al., A Sea-Level Curve for the Canadian Beaufort Shelf, in Can. J. Earth Sci., 22,1383-1395, 1985.

Hinzman, L.D. and D.L. Kane, Potential response  of an Arctic watershed during a period of global warming, ,L
Geophvs. Res.. 97. 2811-2820, 1992.

Hofstetter, R.H., Wetlands in the United States, in Ecosystems of the World, vol. 4B. Mires: Swamp. Boo. Fen.
and  Moor, edited by A.J.P. Gore, Elsevier Sci. Publ., N.Y., pp.  201-244, 1983.

Howes, B.L.,  J.W.H. Dacey, and J.M. Teal,  Annual carbon mineralization and belowground production of Spartina
alterniflora in a New England salt marsh. Ecology. 66. 595-605, 1985.

Idso, S.B., Carbon Dioxide and Global Change: Earth in Transition. 292 pp.. Institute for Biospheric Res., Tempe,
AZ,  1989.

IPCC (1990), Climate Change:  The IPCC Impacts Assessment, Chapter 7: Seasonal  Snow Cover, Ice and
Permafrost,  Co-Chairmen: R.B. Street, Canada, P.I. Melnikov,  USSR.

IPCC (1992), Climate Change 1992: The Supplemental Report to the IPCC Scientific Assessment, prepared by
IPCC Working Group I, edited by J.T. Houghton,  B.A. Callander, S.K. Varney.

Keller, M., T.J. Goreau, S.C. Wofsy, W.A. Kaplan, and M.B. McElroy, Production of nitrous oxide and consumption
of methane by forest soils, Geophvs. Res.  Lett.. 10. 1156-1159,  1983.

Keller, M., W.A. Kaplan, and S.C. Wofsy, Emissions of N2O, CH4, and C02 from tropical forest soils, J. Geophvs.
Res.. 91. 11,791-11,802, 1986.
                                                                                          Page RF-4

-------
Keller, M., Biological sources and sinks of methane in tropical habitats and tropical atmospheric chemistry, Ph. D.
dissertation, Princeton Univ. and Natl. Center for Atmospheric Research, 216 pp., 1990.

Keller, M., M.E. Mitre, and R.F. Stallard, Consumption of atmospheric methane in soils of central Panama: Effects
of agricultural development, Global Bioaeochem. Cycles. 4, 21-27, 1990.

Khalil, M.A.K. (1992), personal communication, Oregon Graduate Institute of Science and Technology, Beaverton,
Oregon.

Khalil, M.A.K., R.A. Rasmussen, J.R.J.  French,  and J.A. Holt, The influence of termites on atmospheric trace
gases: CH4, C02, CHCI3, CO, H2, and light hydrocarbons, J. Geophvs. Res.. 95, 3619-3634, 1990.

Khalil, M.A.K., R.A. Rasmussen, and M.J. Shearer, Trends of atmospheric methane during the 1960s and 1970s,
J. Geophvs. Res.. 94- 18,279-18,288, 1989.

King, G.M. and A.P.S. Adamsen, Effects of temperature on methane oxidation in a forest soil and pure cultures
of the methananotroph Methvlomonas rubra. Appl.  Environ.  Microbiol.. , 1992.

King, G.M., P. Roslev,  and A.P. Adamsen, Controls of  methane oxidation in a Canadian wetland and forest soils,
EOS. 72, 79-80, 1991.

King, G.M. and W.J. Wiebe, Methane release from soils of a Georgia salt marsh, Geochim. Cosmochim.  Acta. 42.
343-348, 1978.

King, S.L., P.O.  Quay, and J.M. Lansdown, The 13C/12C kinetic isotope effect for soil oxidation of methane at
ambient atmospheric concentrations, J.  Geophvs. Res.. 94.  18,273-18,277, 1989.

Kvenvolden, K.A. and M.A. McMenamin (1980), Hydrates of Natural Gas: A Review of Their Geologic Occurrence,
Geological Survey Circular  825 (Washington, D.C.:  US Department of the Interior).

Kvenvolden, K.A., Methane Hydrates and Global Climate, in Global Biogeochemical Cycles, 2:221-229, 1988.

Kvenvolden, K.A. (1990a), Estimate of Current Methane Release from  Gas Hydrates:  1990 Abstracts with
Programs, Geological Society of America, vol. 22, n.  7 p. A195.

Kvenvolden, K.A. (1990b), Arctic Gas Hydrates as a Source of Methane in Global Change: Abstracts, International
Conference on the Role of the Polar Regions in Global  Change, Fairbanks, Alaska, June 11-15, 1990, University
of Alaska Fairbanks, p.211.

Kvenvolden, K.A.,  T.D. Lorenson, T.C. Collett  (1991 a), Arctic Shelf Gas Hydrates as a Possible Source of
Methane:  1991  Abstracts with Programs, Geological  Society of America, v. 23, n. 5, p. A238.

Kvenvolden, K.A., T.S. Collett (1991b),  Permafrost and Gas  Hydrates as Sources of Methane at High Latitudes:
Abstracts, 10th International Symposium on Environmental Biogeochemistry, San Francisco, August  1991, p.22.

Kvenvolden, K.A..Collett,  T.S., and Williams R.S. (1991c), Methane in Permafrost Ice Near Fairbanks,  Alaska:
Trans. Am. Geophysical Union, v. 72, n. 17, p.  67.

Kvenvolden, K.A., and Lorenson, T.D. (1991d), Varying Amounts of Methane in Shallow Permafrost Cores from
Alaska:  Transactions, American Geophysical Union, v. 72, n. 44, Supplement, p. 162.

Kvenvolden, K.A. (1991e), A Review of Arctic Gas  Hydrates as a Source of Methane in Global Change, in
Proceedings of the International Conference on the Role of the Polar Regions in Global Change, ed. G.  Weller, C.
Wilson,  B. Severin, Geophysical Institute, University of Alaska Fairbanks, December  1991.

Kvenvolden, K.A., Lorenson, T.D., and Reeburgh, W.S. (1992), Methane in Permafrost-Preliminary Studies at the
CRREL Permafrost Tunnel near Fox, Alaska:  Transactions, American Geophysical Union, submitted.
                                                                                           Page RF-5

-------
Lachenbruch, A. and Marshall, B., Changing Climate: Geothermal Evidence from Permafrost in the Alaskan Arctic,
in Science 234, 689-696, 1986.

Lafleur, P.M. and W.R. Rouse, The influence of surface cover and climate on energy partitioning and evaporation
in a subarctic wetland. Bound. Layer Meteorol.. 44. 327-347, 1988.

Lafleur, P.M., Evapotranspiration from sedge-dominated wetland surfaces. Aquatic Botany. 37. 341-353, 1990.

Lamborg, M.R. and R.W.F. Hardy, Microbial effects, in C02 and plants: the response of plants to rising levels of
atmospheric carbon dioxide, edited by E.R. Lernon, American Association for the Advancement of Science Selected
Symposia Series, Westview  Press, Boulder, CO, 1983.

Lashof,  D.A.,  The dynamic greenhouse: Feedback  processes  that may influence future  concentrations of
atmospheric trace gases and climatic change, Climatic Change. 14. 213-242, 1989.

Livingston,  G.P. and L.A. Morrissey, Methane emissions from Alaska arctic tundra in response to climatic change,
in: Proc. Int. Conf. on the Role of the Polar Regions in Global Change, in press, 1991.

Lorius, C., et al.. Nature, v.  347, p. 139, 1990.

MacDonald, G.J., Role of Methane Clathrates in Past and Future Climates, in Climatic Change, 16, 247-281,1990.

MacDonald, G.J. (1983), The Many Origins of Natural Gas, in Journal of Petroleum Geology, v. 5, n. 4, pp. 341-
362.

Martens, C.S., C.A. Kelley,  J.P. Chanton, arid W. Showers,  Carbon and hydrogen isotopic characterization of
methane from wetlands and lakes of the Yukon-Kuskokwim Delta, J. Geophvs. Res.. 97, 16689-166701,1992.

Matthews,  E., Global vegetation and land use: New high-resolution data bases for climate studies, J. Climate Appl.
Meteorol.. 22, 474-487, 1983.

Matthews,  E. and I. Fung, Methane emission from natural wetlands: Global distribution, area, and environmental
characteristics of sources. Global Bioaeochem. Cycles. 1_, 61-86, 1987.

Mclver, R.D., Gas Hydrates,  in Long-Term Energy Resources, vol. 1, edited by R.F. Meyer and J.C. Olson, pp. 713-
726, Pitman, Boston, MA.,  1981.

Meyer, R.F., Speculations on Oil and Gas Resources in Small Fields and Unconventional Deposits, in Long-Term
Energy Resources, vol. 1, edited by R.F. Meyer and J.C. Olson,  pp. 49-72, Pitman, Boston, MA., 1981.

Miller, S.R. (1974), The Nature and Occurrence of Clathrate Hydrates, in Natural Gases in Marine Sediments. I.R.
Kaplan, ed. Plenum, New York, pp. 151-170.

Mitchell, J.F.B., S. Manabe, T. Tokioka, and V. Meleshko, Equilibrium climate change, in: Climate Change: The
IPCC Scientific Assessment, edited by J.T. Houghton, G.J. Jenkins, and J.J. Ephraums, Cambridge Univ. Press,
N.Y.,  pp. 131-172, 1990.

Mitchell, J.F.B., The greenhouse effect and climate change. Rev, of Geophvs.. 24. 115-139, 1989.

Moore, T.R., N. Roulet, and  R. Knowles, Spatial and temporal variations of methane flux from  subarctic/northern
boreal fens, Global Biogeochem. Cycles. 4, 29-46, 1990.

Moore, T.R. and R. Knowles, Methane and carbon dioxide evolution from subarctic fens, Can. J. Soil Sci.. 67, 77-
81,1987.

Moore, T.R. and R. Knowles, Methane emissions from fen, bog, and swamp peatlands in Quebec, Biogeochem..
11,45-61, 1990.
                                                                                           Page RF-6

-------
Moore, T.R., A. Heyes, S. Holland, W.R. Rouse, N.T. Roulet, and L. Klinger, Spatial and temporal variations of
methane emissions in the Hudson Bay Lowlands, EOS.  72. 84, 1991.

Moraes, F. and M.A.K. Khalil, The effect of permafrost on atmospheric concentrations of CH4, CO2, H2, and CO,
EOS. 73, 62, 1992.

Morrissey, L.A. and G.P. Livingston, Methane emissions  from Alaska arctic tundra: An assessment of local spatial
variability, J. Geophvs. Res.. 97, 16661-16670, 1992.

Mosier, A., D. Schimel, D. Valentine,  K. Bronson, and  W. Parton, Methane and nitrous  oxide fluxes in  native,
fertilized and cultivated grasslands, Nature. 350. 330-332, 1991.

Mulkey, Lee, personal communication, US EPA Environmental Research Laboratory, 960 College Station Road,
Athens, Georgia  30613-0801,1993.

Naiman, R.J., T. Manning, and C.A. Johnston, Beaver populations and tropospheric methane emissions in boreal
wetlands,  Bioaeochem.. 12. 1-15, 1991.

Nisbet, E.G., Some Northern Sources of Atmospheric Methane: Production, History and Future Implications, in
Canadian Journal of Earth Science, vol 26, pp. 1603-1611, 1989.

Nisbet, E.G. (1990), Did the Release of Methane from Hydrate Accelerate the End of the Last Ice Age?,  in Can.
J. Earth Sci., 1989.

Nisbet, E.G. (1992), personal communication,  Department of Geological Sciences, University of Saskatchewan,
Saskatoon, Saskatchewan, Canada.

Nisbet, E.G. and B. Ingham, Methane output from natural and quasi-natural sources in the subsurface, sould,
bogs, and shallow wetlands and shallow arctic seas: the potential  for change and biotic and abiotic feedback,
submitted. Department of Geology, Royal Holloway, University of London., London TW20 OEX, U.K.

Nuttle, W.K. and H.F. Hemond, Salt marsh hydrology: Implications for biogeochemical fluxes to the atmosphere
and estuaries, Global Biogeochem. Cycles. 2,  91-114,  1988.

Osterkamp, T.E., and Lachenbruch, A.H., Thermal Regime of Permafrost in Alaska and Predicted Global Warming,
in Journal of Cold Regions  Engineering, v. 4, n.  1, pp.  38-42, March, 1990.

Osterkamp, T.E., T. Fei (1992), Potential Occurence of Permafrost and Gas Hydrates in the Continental Shelf Near
Lonely, Alaska, submitted for publication in the Proceedings of the Sixth Int. Conf. on Permafrost, Beijing, China,
July,  1993.

Osterkamp, T.E., Response of Alaskan Permafrost to Climate, Proceedings of the Fourth Int. Conf. on Permafrost,
July 18-23, 1983, Fairbanks,  AK, Nat. Acad.  of Sci., Washinton, D.C.

Post, W.M. (ed.), Report of a  Workshop on Climate Feedbacks and the Role of Peatlands, Tundra, and Boreal
Ecosystems in the Global Carbon Cycle, Oak Ridge Nat'l Lab.,  Env. Sci. Div. Publ. #3289, 1990.

Pulliam, W.M., Carbon dioxide and methane exports from a southeastern floodplain swamp, Ecol. Monoar.. in
press, 1992.

Pulliam, W.M. and J.L. Meyer, Methane emissions from floodplain swamps of the Ogeechee River: Long-term
patterns and effects of climate change, Bioaeochem.. in press, 1992.

Raynaud,  D., et al., Science, vol 259, pp. 926 - 934, 1993.

Reeburgh, W.S. and S.C. Whalen, High latitude ecosystems as CH4 sources, in: Trace Gas Exchange in a Global
Perspective, edited by B. Svensson and D. Ojima, Munksgaard, Copenhagen, SCOPE/IGBP, in press.
                                                                                           Page RF-7

-------
Revelle, R.R., Methane Hydrate in Continental Slope Sediments and Increasing Atmospheric Carbon Dioxide, in
Changing Climates, pp. 252-261, National Academy Press, Washington D.C., 1983.

Ritter, J.A., J.D.W. Barrick, G.W. Sachse, G.L. Gregory, M.A. Woerner, C.E. Watson, G.F. Hill, and J.E. Collins,
Airborne flux measurements of trace species in an arctic boundary layer, J. Geophvs. Res.. 97, 16601-16625,
1992.

Ritter, J.A.,  C. Watson, J. Barrick, G. Sachse, J. Collins, G. Gregory, B. Anderson, and  M. Woerner, Airborne
boundary-layer measurements of heat, moisture, CH4, CO, and O3 fluxes over Canadian boreal forest and northern
wetland regions, EOS. 72. 84, 1991.

Rodhe, H., A comparison of the contribution of various gases to the greenhouse effect, Science. 248.1217-1219,
1990.

Roulet, N.T., R. Ash, and T.R. Moore,  Low boreal wetlands as a source of atmospheric methane, J. Geophvs. Res..
97, 3739-3749, 1992a.

Roulet, N., T. Moore, J. Bubier, and P. LaFleur, Northern fens: Methane flux and climatic change, Tellus. 44B. 100-
105, 1992.

Sass, R.L., P.M. Fisher, F.T. Turner, and M.F. Jund, Methane emission from rice fields as influenced by solar
radiation, temperature, and straw incorporation. Global Bioaeochem. Cycles. 5, 335-350, 1991.

Scharffe, D., W.M. Hao, L. Donoso, P.J. Crutzen, and E. Sanhueza, Soil fluxes and atmospheric concentration of
CO and CH4 in the northern part of the Guayana Shield, Venezuela, J. Geophvs. Res.. 95. 22,475-22,480,1990.

Schiff, H.I.,  D.R. Karecki, F.J. Lubkin,  R. Big, and G.I.  Mackay, A tunable diode laser system for CH4 flux
measurements from a small aircraft,  EOS, 72,  78, 1991.

Schutz, H., A. Holzapfel-Pschorn, R. Conrad, H. Rennenberg, and W. Seiler, A three year continuous record on the
influence of daytime, season, and fertilizer treatment on methane emission rates from an Italian rice paddy, J^
Geophvs. Res.. 94, 16,405-16,416, 1989.

Schutz, H., P. Schroder, and H. Rennenberg, Role of plants in regulating the methane flux to the atmosphere, in
Environmental  and Metabolic Controls on Trace Gas Emissions from Plants, edited by T. Sharkey, E. Holland, and
H. Mooney,  Academic Press, New York, pp. 29-63, 1991.

Sebacher, D.I., R.C. Harriss, and K.B. Bartlett,  Methane emissions to the atmosphere through aquatic plants, J^
Environ. Qua I..  14. 40-46, 1985.

Sebacher, D.I., R.C. Harriss, K.B. Bartlett, S.M. Sebacher, and S.S. Grice, Atmospheric methane sources: Alaskan
tundra bogs, an alpine fen, and a subarctic boreal marsh, Tellus. 38B. 1-10, 1986.

Seiler, W., R. Conrad, and D. Scharffe, Field studies of methane emission from termite nests  into the atmosphere
and measurement of methane uptake by tropical soils, J. Atmos. Chem.. 1, 171-186, 1984.

Smith, L.K. and W.M. Lewis, Methane emissions from the Orinoco  River floodplain, Venezuela,  ASLO 92, Aquatic
Sci. Mtg., Santa Fe, NM, Feb. 9-14, 1992.

Smith, Michael, (1992), personal communication, Carlton University, Ottawa, Ontario, Canada.

Stauffer, B., G. Fischer, A. Neftel, and H. Oeschger, Increase of atmospheric methane recorded in Antarctic ice
core, Science. 229. 1386-1388, 1985.

Steudler, P.A., R.D. Bowden, J.M. Melillo, and J.D.  Aber, Influence of nitrogen fertilization on methane uptake in
temperate forest soils, Nature. 341.  314-316, 1989.
                                                                                           Page RF-8

-------
Striegl,  R.G., T.A. McConnaughey,  D.C. Thorstenson,  E.P. Weeks, and  J.C. Woodward,  Consumption  of
atmospheric methane by desert soils, Nature.  357. 145-147, 1992.

Svensson, B.H., Methane production in tundra peat, in: Microbiai Production and Utilization of Gases (H2. CH^.
CO), edited by H.G. Schlegel, G. Gottschalk, and N. Pfennig, E. Goltze, Gottingen, pp.  135-139, 1976.

Svensson, B.H. and T. Ross wall, In situ methane production from acid peat in plant communities with different
moisture regimes in a subarctic mire, Oikos. 43. 341-350, 1984.

lathy, J.P., B. Cros, R.A. Delmas, A. Marenco, J. Servant, and M. Labat, Methane emission from flooded forest
in Central Africa, J. Geophvs. Res..  9_7, 6159-6168, 1992.

Trivett, N.B.A., D. Worthy, and K. Bruce,  Surface measurements of carbon dioxide and methane at Alert during
an Arctic haze event in April, 1986, J. Atmos. Chem.. 9, 383-397, 1989.

Vitt, D., S. Bayley, T. Jin. L. Halsey, B.  Parker, and R.  Craik,  Methane and  Carbon Dioxide Production from
Wetlands in Boreal Alberta, rept. to  Alberta Environment, contract #90-0270, 39 pp., 1990.

Wassmann, R., U.G. Thein, M.J. Whiticar, H. Rennenberg, W. Seller, and W.J. Junk, Methane emissions from the
Amazon floodplain: Characterization of production and transport. Global Bioaeochem. Cycles. 6_,  3-13, 1992.

Whalen, S.C. and W.S. Reeburgh, Interannual  variations in tundra methane emissions: A four-year time series at
fixed sites, Global Bioqeochem. Cycles. 6, 139-159, 1992.

Whalen, S.C., W.S. Reeburgh, and K.S. Kizer,  Methane consumption and emission by taiga, Global Bioaeochem.
Cycles.  5, 261-273, 1991.

Whalen, S.C. and W.S. Reeburgh, Consumption of atmospheric methane by tundra soils, Nature. 346. 160-162,
1990a.

Whalen, S.C. and W.S. Reeburgh, A methane flux transect along the trans-Alaska pipeline haul road, Tellus. 42B.
237-249, 1990b.

Whalen, S.C., W.S. Reeburgh, and K.A. Sandbeck, Rapid methane oxidation in a landfill cover soil, Appl. Environ.
Microbiol.. 56, 3405-3411, 1990c.

Whalen, S.C. and W.S. Reeburgh, A methane flux time series for tundra environments, Global Bioaeochem. Cycles.
2, 399-409, 1988.

Whiting, G.J., J.P. Chanton, D.S. Bartlett, and J.D. Happell, Relationships between CH4 emission, biomass, and
C02 exchange in a subtropical grassland, J. Geophvs. Res.. 96. 13,067-13,071, 1991.

Whiting,  G.J. and J.P.  Chanton, Plant  dependent methane emissions in a subarctic  Canadian  fen, Global
Bioqeochem.  Cycles.  6, 225-231, 1992.

Wilson, J.O., P.M. Crill, K.B. Bartlett, D.I. Sebacher, R.C. Harriss, and R.L. Sass, Seasonal variation of methane
emissions from a temperate swamp, Bioaeochem.. 8, 55-71, 1989.

Yarrington, M.R.  and  D.D. Wynn-Williams, Methanogenesis and the anaerobic microbiology of a wet moss
community at Signy Island, in: Antarctic Nutrient Cycles and Food Webs, edited by W.R. Siegfried, P.R. Condy,
and R.M. Laws, Springer-Verlag, Berlin, pp. 134-139, 1985.

Yavitt, J.B., G.E. Lang, and A.J. Sexstone, Methane fluxes in wetland and forest soils, beaver ponds, and low-
order streams of a temperate forest ecosystem, J. Geophvs. Res.. 95. 22,463-22,474, 1990.

Zehnder, A.J.B., Ecology of methane formation, in: Water Pollution Microbiology, vol. 2, edited by R. Mitchell, J.
Wiley and Sons, New York, pp. 349-376, 1978.
                                                                                           Page RF-9

-------
                                         APPENDICES
Appendix A: List of UNH Workshop Participants
 Karen Bartlett
 Complex Systems Res. Center
 Institute for the Study of
 Earth, Oceans, and Space
 Univ. of New Hampshire
 Durham, NH 03824

 Jeffrey Chanton
 Dept. of Oceanography
 Florida State Univ.
 Tallahassee, FL 32306


 Patrick Crill
 Complex Systems Res. Center
 Institute for the Study of
 Earth, Oceans, and Space
 Univ. of New Hampshire
 Durham, NH 03824

 Stephen Frolking
 Complex Systems Res. Center
 Institute for the Study of
 Earth, Oceans, and Space
 Univ. of New Hampshire
 Durham, NH 03824

 Robert Harriss
 Complex Systems Res. Center
 Institute for the Study of
 Earth, Oceans, and Space
 Univ. of New Hampshire
 Durham, NH  03824

 Harold Hemond
 48-419
 Massachusetts Inst. of Technology
 Cambridge, MA  02319

 Kathleen Hogan
 Global Change Div. ANR-445
 U.S. E.P.A.
 401 M Street, S.W.
 Washington, DC  20460

 Michael Keller
 USDA Forest Service
 Institute of Tropical Forestry
 Call Box 25000
 Rio Piedras, Puerto Rico 00928
Daniel Lashof
Natural Resources Defense Council
1350 New York Ave., N.W.
Washington, DC 20005
Elaine Matthews
NASA Goddard Space Flight Center
Institute for Space Studies
2880 Broadway
New York, NY  10025

William Pulliam
Jones Ecological Res. Center
Ichauway, Rt. 2, Box 2324
Newton, GA 31770
Jeffrey Ross
Bruce Company
501 3rd St. N.W. Suite 260
Washington, DC  20001
Nigel Roulet
Dept. of Geography
York Univ.
4700 Keele St.
North York, Ontario CANADA
M3JIP3

Margaret Torn
Energy and Resources Group
Univ. of California, Berkeley
Berkeley, CA 94720

Steven Whalen
Inst. of Marine Science
School of Fisheries and Ocean Sciences
Univ. of Alaska
Fairbanks, AK 99775-1080
                                                                                        Page AP-1

-------
Appendix B: Methane Flux from Tropical Wetlands
  (Fluxes made using enclosure techniques unless
   otherwise noted; Flux units are mgCH4/m2/d)
HABITAT


Open water
Non-forest swamps
Flooded forests

Open water
Non-forest swamps
Flooded forests

Open water
Non-forest swamps
Flooded forests

Open water
Non-forest swamps
Flooded forests

Open water
Non-forest swamps
Flooded forests
Open water
Non-forest swamps
Flooded forests
Lakes (w/
macrophytes)-
Open water
Non-forest swamps
Flooded forests

Lakes (w/o
macrophytes)-
Open water
Flooded forests


Lake 1 (w/
macrophytes)-
Open water
Non-forest swamps
Flooded forests
LATITUDE


3
3
3

3
3
3

3
3
3

3
3
3

3
3
3
3
3
3


9
9
9



9
9




9
9
9
AV6.
FLUX

881
3901
751

27
230
192

74
201
126

40
131
7.1

44
214
150
36
35
75


7.5
29.6
257.6



73.8
307.2




22.6
61.3
174.4
ANNUAL
FLUX
(g/m2/yr)
_
-
-

.
-
-

.
-
-

.
-
-

162
782
552
132
-
-


-
-
-



-
-




-
-
-
N


36
25
16

41
90
55

116
85
58

40
31
11

90
63
31
.
-
-


26
16
18



14
8




30
37
24
RANGE


0.5-665
0.8-1976
1-533

-10.5-111
0-1224
0-2997

0-1 1 60
-11.3-1600
0-840

.
-
-

0-350
0-2600
0-2700
.
-
-


0-48
0-132
0-1872



2-530
3-2288




1-159
-4-918
0-1 648
SITE


Amazon
flood
plain

Central
Amazon
flood-
plain
Amazon
flood-
plain

Amazon
flood
plain

Lakes
near
Manaus
Lake
near
Manaus


Orinoco
flood-
plain



Orinoco
flood-
plain



Orinoco
flood-
plain
MEAS.
PERIOD

July
(falling
high
water)
July-
Sept.
(high
water)
Apr.-May
(rising
high
water)
Nov.-
Dec.
(low
water)
Annual


Annual
Feb-Aug
Feb-Aug


July-
Oct.
(high
water)


July-
Oct.
(high
water)


July-
Dec.

REF
.

1



2



3



4






5




6














                                                         Page AP-2

-------
HABITAT


Lake 2 (w/o
macrophytes)-
Open water
Flooded forest

Flooded forest-
Flooded soils
Wet/moist
soils
Dry soils
Open water-
Shallow (0.5-2m)
Middle (4-6m)
Deep (7-1 Om)
Swamp-
Well-drained
(rain forest)
Marsh
Swamp forest
LATITUDE




9
9


1
1

1

9
9
9


9
9
9
AVG.
FLUX



32.6
248


106
4.9

-1.9

9673
3953
543


0.71
379
346
ANNUAL
FLUX
(g/m2/yr)


-
-


-
-

-

.
.
-


0.262
1382
1262
N




57
19


11
4

5

15*
154
64


144
36
90
RANGE




0-587
0-2736


9.9-550
1.2-7.6

-0.8-4.6

297-3635
127-1148
0-207


-2-13
0-2600
0-2200
SITE




Orinoco
flood-
plain

Congo
River
basin


Gatun
Lake,
Panama


Mojinga
Swamp,
Panama
MEAS.
PERIOD



July-
Dec.


Feb. &
Oct.



Feb.-
Oct.



Annual


REF








7




8




8


1 Corrected from original data due to 37% error in bubble flux.  See Bartlett et al. (1990).

2 Annual flux is average flux multiplied by 365 days.

3 Bubble flux only; more than 97% of total flux.

4 24-hour flux measurements rather than short-term (15-30 minutes).

REFERENCES:

        1) Devoletal., 1988
        2) Bartlett et al., 1988
        3) Bartlett et al., 1990
        4) Devol etal., 1990
        5) Wassmann et al.,  1992
        6) Smith and Lewis,  1992; Smith (pers. com.)
        7) Tathy et al., 1992
        8) Keller, 1990
                                                                                              Page AP-3

-------
Appendix C: Methane Flux from Temperate and Subtropical Wetlands
          (Fluxes made using enclosure techniques unless
           otherwise noted; Flux units are mgCH4/m2/d)
HABITAT
Salt marsh-
short S.alter-
niflora
tall S.alter-
niflora
Salt marsh-
short S.alter-
niflora
tall S.alter-
niflora
salt meadow
' short S.alter-
niflora
short S.alter-
niflora
tall S.alter-
niflgra
salt meadow
high marsh
tall S. alter-
niflora
short S. alter-
niflora
Juncus roemer-
ianus
short S. alter-
niflora
Salt marsh-
short S. alter-
niflora
Graminoid
marshes-
Saline
LATITUDE
31
-
37
-
-
39
38
"
-
33
-
38
30
-
41
30
AVG.
FLUX
145
15.87
1.2
3.0
5.0
2.0
0.5
-0.8
1.5
-1.9
1.5
0.4
13.4
3.9
0.6
4.51
15.7
ANNUAL
FLUX
(g/m2/yr)
53.1
0.4
1.3
1.2
0.43
-
-
-
-
-
-
-
-
-
1.6
5.7
N
17
19
24
29
47
6
4
3
2
25
8
3
9
2
24
36
RANGE
0.24-
1920
0.02-
144
0-13.9
-1.1-
16.2
-2.7-
21.3
-1.0-
3.9
-1.7-0
1.2-
1.7
-1.8-
-2.0
-3.9-
9.6
0-4.8
2.2-19.2
-3.2-6.3
0-1.3
1.3-12
1 .3-48
SITE
Sapelo
Isl.,
GA
-
Bay
Tree,
VA
"
n
Lewes,
DE
Wallops
Isl., VA
ti
-
George-
town, SC
"
Sapelo
Isl., GA
Panacea,
FL

Sippiwis-
sett, MA
Miss.
Delta,
LA
MEAS.
PERIOD
annual

annual


June
June


Apr &
Aug

Nov
Sept-
Dec

annual
annual
REF.
1

2











3
4
                                                                 Page AP-4

-------
HABITAT
Brackish
Fresh
Open water-
Saline
Brackish
Fresh
Graminoid marsh-
Saline
Brackish
Brackish/Fresh
Ombrotrophic
bog
Minerotrophic
fen (5 sites)
Beaver pond
Running stream
Minerotrophic
fen-1 989-1 990
1990-1991
Forested swamp-
swamp bank
Peltandra
Smartweed
Ash tree
Swamp
Farm ponds
Maple/gum
swamp
Cypress swamp-
floodplain
LATITUDE
"
-
n
R
n
37
-
-
39
n
-
n
43
43
37
37
n
n
42
-
36
33
AVG.
FLUX
267
587
4.8
17
49
16
64.6
53.5
-1.14
3.56
250
300
133
291
117
155
83
152
3562
4402
-
9.9
ANNUAL
FLUX
(g/m2/yr)
97
213
1.7
6.2
18.2
5.6
22.4
18.2
.
-
-
-
49
107
43.7
41.7
-
n
-
-
0.5
-
N
36
27
13
13
13
21
21
21
105
527
105
105
38
107
29
29
29
29
7
25
37
6
RANGE
13-1300
0-2600
.
-
-
1-46
1.3-180
-3-259
-6.7-
29.6
-7.7-748
1-1400
2-4000
7-941
5-3563
0-475
4-469
0-405
0-1005
120-160
92-1100
-6-20
4.6-21.8
SITE
-
Tt
.
-
-
Queen's
Creek,
VA
n
tt
Buckle's
Bog, MD
Big Run
Bog, WV
ft
N
Sallie's
Fen, NH
-
Newport
News
Swamp,
VA
-
ri
N
S.E.
Mich.
N
Dismal
Swamp,
VA
Four Holes
Swamp,
SC
MEAS.
PERIOD





annual


annual



annual

annual



May-
Sept

annual
June
REF.





5


6



7

8



9

10
11
Page AP-5

-------
HABITAT
deep-water
flowing water
Bottomland
hardwoods &
gum - cypress
swamps
Shrub swamp
Aquatics/Prairie
Cypress swamp
Gum/Bay swamp
Lakes
Wet prairie/
Sawgrass
Wetland forest
Salt water
mangroves
Impoundments/di
sturbed wetlands
Hardwood
hammock
Red mangroves
Dwarf cypress/
sawgrass
Spikerush
Sawgrass <1m
LATITUDE
31
26
32
31
31
-
n
"
26
n
-
n
26
if
n
n
-
AVG.
FLUX
92.3
67.0
-
149
130.
4
39.8
69.6
115.
9
61
59
4
74
0
4.2
7.5
29.4
38.8
ANNUAL
FLUX
(g/m2/yr)
-
-
10-34
-
-
-
-
-
-
-
-
-
-
-
-
-
-
N
3
6
-
42
50
47
14
7
122
22
17
32
3
17
25
7
62
RANGE
10-256
8.2-265
-10-361
-7.5-
1250
-7.9-
1000
-10-442
-7.5-293
30.8-217
0-624
-3-274
1.9-7.7
5.9-198
-
-
-
-
-
SITE
Okefen-
okee
Swamp,
GA
Cork-
screw
Swamp,
FL
Ogeechee
River,
GA
Okefen-
okee
Swamp,
GA
Okefen-
okee
Swamp,
GA
-
H
-
Ever-
glades,
FL

n

Ever-
glades,
FL
n
n
-
-
MEAS.
PERIOD


annual
(multi-
yr)
annual
annual



annual



Dec-
Feb




REF.


12
13




14



15




Page AP-6

-------
HABITAT
Sawgrass/Spike-
rush/Periphyton
Swamp forest
Sawgrass >1m
Dwarf red
mangrove
Sawgrass/Muhly
grass/Peri-
phyton
Tall saw grass/
Muhly grass
Mangroves
Forested swamp
Subtropical
estuary
Sawgrass
Pond open water
LATITUDE

it
-
?t
26
-
-
28
n
26
-
AVG.
FLUX
45.1
68.9
71.9
81.9
883
7.93
823
1133
2.63
107
624
ANNUAL
FLUX
(g/m2/yr)
-
-
-

-
-
-
-
-
-
-
N
8
13
24
19
6
7
4
12
80
60
9
RANGE
-
-
-
-
-
-
-
-
-
9-2387
11-
2646
SITE
-
tt
-
-
Ever-
glades,
FL
n
n
Tampa
Bay, FL
n
Ever-
glades,
FL
-
MEAS.
PERIOD




annual




annual

REF.




16




17

1 Measured from cores.
2
  Bubble flux only.
3 Diffusive flux only, calculated from surface water concentration data.
REFERENCES:
        1) King and Wiebe, 1978
        2) Bartlett et al., 1985
        3) Howes et al., 1985
        4) DeLaune et al., 1983
        5) Bartlett et al., 1987
        6) Yavitt et al.,  1990
        7) Grill, unpublished data
        8) Wilson et al., 1989
9) Baker-Blocker etal., 1977
10) Harriss etal., 1982
11) Harriss and Sebacher,  1981
12) Pulliam, in press; Pulliam & Meyer, in press
13) Bartlett, unpublished data
14) Harriss et al., 1988
15) Bartlett et al., 1989
16) Barber etal., 1988
17) Burke etal., 1988
                                                                                               Page AP-7

-------
Appendix D: Methane Flux Measurements from Northern Wetlands
         (Fluxes made using enclosure techniques unless
           otherwise noted; Flux units are CH4/m2/d)
                          ARCTIC
HABITAT
Subarctic mire:
ombrotrophic
intermediate
minerotrophic
Wet coastal tundra
Moist tundra
Meadow tundra
Alpine fen
Boreal marsh
Tussock tundra
composite (1987)
Wet meadow
composite (1987)
Moss (1987)
Moss (1988)
Moss (1989)
Moss (1990)
Intertussock
(1987)
Intertussock
(1988)
Intertussock
(1989)
Intertussock
(1990)
LATITUDE
68
-
-
63-70
"
-
n
n
65
-

65
-

n
n
-
-
AVG.
FLUX
11.6
58
360
119
4.9
40
289
106
22.4
32.2
O.9
10.1
29
3.8
2.8
25.2
12.4
5.9
ANNUAL
FLUX
(g/m2/yr)
0.13
1.44
30.5
-
-
-
-
-
-
-
0.5
4.4
4.8
0.5
0.6
3.9
4.3
0.8
N
95
25
20
44
12
14
6
7
-
-
46
57
58
51
39
54
57
51
RANGE
0.3-
29
8.6-
112
80-
950
34-266
0.3-12
9-77
244-344
86-122
-
-
0-17
0-146
0-367
0-26
0-34
0-292
0-145
0-57
SITE
Stordalen,
Sweden


Alaska
North
Slope &
Denali




Alaska,
Fairbanks









MEAS.
PERIOD
June-
Sept.


Aug.




Annual









REF.
1,2


3




4,5









                                                               Page AP-8

-------
HABITAT
Carex sedge
(1987)
Car ax sedge
(1988)
Car ax sedge
(1989)
Carex sedge
(1990)
Eriophorum
tussocks (1987)
Tussocks (1988)
Tussocks (1989)
Tussocks (1990)
Lakes & ponds
Alpine tundra
Moist tundra
Wet tundra
Low brush-Muskeg
bog
Spruce forest
Meadow & tussock
tundras
Tussock tundra:
tussocks
intertussocks
Meadow tundra
Lakes & ponds
High centered
polygons
Low centered
polygons:
troughs
basins
LATITUDE
n
n
"
it
-
-
-
n
66-70
-
n
n

n
68-70
69
-
-
"
-

-
AVG.
FLUX
33.5
3.2
23
451.9
29.5
60.8
48.8
96.8
21
0.6
31
90
45
4.6
30
3.4
2.9
64.4
37.5
4.9
61.9
46.1
ANNUAL
FLUX
(g/m2/yr)
4.9
0.8
4.3
60.6
8
11.4
8.1
13.6
-
-
-
-
-
-
-
-
-
-
-
-
-
-
N
33
57
60
51
48
57
60
51
6
8
42
18
6
18
27
6
8
23
31
5
10
24
RANGE
<1-105
0-12
0-104
0-2216
0-167
0-653
0-445
0-302
4.6-131
-0.2-6.3
0-159
0-265
12-101
-0.3-67
-1-145
-
-
-
-
-
-
-
SITE








Alaska,
North
Slope





Alaska,
N. Slope
& Foot-hills
Alaska,
N. Slope






MEAS.
PERIOD








July-Aug.





Aug.
Aug.






REF.








6





7
8






Page AP-9

-------
HABITAT
rims
Wet tundra
Very wet tundra
(Calc. regional
mean, 1987)+
Wet meadow
tundra
Upland tundra
Large lakes*
Small lakes*
Lake vegetation
(Calc. regional
mean)"1"
Small lakes-
bubbles(calc.)
Dry tundra
Wet meadow
tundra
Lake (wind at 5
m/sec)
(integrated areal
mean)
(integrated areal
means)
Wet moss carpet
LATITUDE

69
n
n
61
n
-
-
-
-
61
61
n
-
-
61
~70S
AVG.
FLUX
12.1
100.1
253.9
17.3
144
2.3
3.8
77
89
48.5
112
11
29
57
25
44
1.6
ANNUAL
FLUX
(g/m2/yr)
-
-
-

-
-
-
-
-

-
-
-
-
-
-
-
N
15



73
82
12
32
51

-
eddy
corr.



eddy
corr.
208
RANGE
-
-
-

16-426
-2.1-18
-
-
63-154

-
-
-
-
-
25-85
0.03-
20.3
SITE

Alaska,
N.SIope


Alaska,
Yukon
Kuskokwim
Delta





Alaska
Y-K Delta
Alaska,
Y-K Delta



Alaska,
Y-K Delta
Antarctica,
Signy Isl.
MEAS.
PERIOD

Aug.


July-Aug.





July-Aug.
July-Aug.



July-Aug.
Dec. -Mar.
REF.

9


10





11
12



13
14
Page AP-10

-------
BOREAL
HABITAT
Forested bogs
Nonforested bogs
Forested fens
Wild rice bed
Sedge meadow
Forested bogs
Nonforested bogs
Forested fen
Nonforested fens
Nonforested bog
Forested bog
(hummock)
Forested bog
(hollow)
Fen lagg
Nonforested bog
Nonforested fen
Nonforested bog
Nonforested poor
fen
Nonforested rich
fen
Forested rich fen
Sedge meadow
LATITUTE
47
-
"
-
n
47
H
-
"
47
47
n
"
n
n
54-55
n
-
n
n
AVG.
FLUX
100
306
85
493
664
89
199
142
348
177
21
93
121
356
402
0
1
65
24
148
ANNUAL
FLUX
(g/m2/yr)
-
-
-
-
-
-
-
-
-
-
3.5
13.8
12.6
43.1
65.7
0
0.1
9.5
3.4
21.7
N
8
24
5
4
1
55
77
12
35
eddy
corr.
36
36
27
68
37
90
90
90
90
90
RANGE
1 9-206
33-1943
3-171
127-883
-
1 1 -694
1 8-866
68-263
152-711
120-270
2-48
6-246
-1-482
0-1056
11-767
0
0-22
0-1976
0-1820
0-1985
SITE
Minnesota,
Marcell
Forest &
Zerkel




Minnesota,
Marcell
Forest &
Red Lake



Minnesota,
Marcell
Forest
Minnesota,
Marcell
Forest




Alberta,
Canada




MEAS.
PERIOD
Aug.




May-
Aug.



Aug.
Annual




May -
Oct.




REF
15




16



17
18




19




                                 Page AP-11

-------
HABITAT
Beaver pond
Nonforested fens
Nonforested fens:
center
margin
pools/
flooded
Forest/fen margin
Patterned
nonforested fens:
ridges
pools
(calc. regional
mean) *
Horiz. rich fen
Horiz. poor fen
Ribbed fen:
ridge
pools
Basin swamps
Domed bog:
center
margin
Beaver ponds
LATITUTE
H
55
55
m
'
55
ii
it
H
55
-
n
n
45
n
n
45
AVG.
FLUX
518
30.5
72
30
33
28
7.5
48
18
-
-
-
-
-
-
-
90.6
29.7
47.4
ANNUAL
FLUX
(g/m2/yr)
78.2
-
-
-
-
-
-
-

3
9.8
1.3
4.5/9.9
1.2/4.2
0.1
0.1
7.6
tt
N
90
80
205
195
185
255
110
85

-
-
-
-
-
-
-
56
65
65
RANGE
0-12,068
0-112
29-125
9-65
24-40
0.6-51
5-9
32-66

-3-176
1 2-343
1-25
1-26O
-3-207
-11-10
-10-9
0.9-246
0.3-300
0.2-369
SITE

Schef-
ferville,
Canada
Schef-
ferville,
Canada






Schef-
ferville,
Canada



Mont St.
Hilaire
Canada


low
boreal
forest,
Canada
MEAS.
PERIOD

June-
Aug.
June-
Sept.






May-
Sept.
(multi-
Yr)



May-
Aug.
(multi-
yr)


May-
Oct.
REF

20
21






22






23
Pafle AP-12

-------
HABITAT
Conifer swamps
Mixed hardwood
swamps
Thicket swamps
Marshes
Open bog
Forested bog
Fen
Blanket bog:
pool
lawn
hummock
(integrated area!
means)
(integrated area!
mean)
Ponds & lakes
Fen ponds
Open water
Marshes
Shrub & Treed Fen
Open Fen
LATITUTE
II
45
-
-
-
-
n
55
n
"
55
55
55
56
55

ti
H
AVG.
FLUX
7.1
0.15
0.2
0.2
1.2
0.25
69.3
0.4
1.2
0.5
20.6
5.8
3.0
-
-
-
-
16
26
160
12
31
2.5
7.9
ANNUAL
FLUX
(g/m2/yr)
0.18
0.1
It
4.7
0.1
n
1.7
tf
0.4
9.3
5.3
1.3
-
-
-
-
1.5
2.3
0.4
0.7
N
123
132
148
149
141
139
145
144
72
134
65
68
62
36
36
36
-
-
-
-
-
-
-
-
RANGE
-0.2-236
0.1-10
-0.2-6
-0.2-9
-0.3-28
-5.8-10
0-304
-0.3-37
-0.1-36
-0.3-26
-0.1-140
-0.1-107
-0.2-78.2
-
-
-
0-50
-
-
-
0.2-146
-2.3-274
-2.4-32
-1.6-298
SITE







Moor
House
Nat. Res
England


Hudson Bay
Low-lands,
Canada
Kino-sheo
Lake,
Canada
Kinosheo
Lake,
Canada
Coastal &
int. Hudson
Bay
South-
ern Hudson
Bay Low-
lands
Canada



MEAS.
PERIOD







Annual


July
July
July

June-
Oct.



REF







24


25,
26
27
28

29



Page AP-13

-------
HABITAT
Fen pools
Bog pools
Open bog
Shrub-rich Bog
Treed Bog
Conifer Forest
LATITUTE
-
it
-
-
"
-
AVG.
FLUX
133
60
54
48
1.8
3.3
ANNUAL
FLUX
(g/m2/yr)
13.8
6.1
4.6
4.0
0.2
0.2
N
-
-
-
-
-
-
RANGE
21-544
2.2-665
-1.7-1356
-1.5-1627
-1.7-66
-2.2-50
SITE






MEAS.
PERIOD






REF






  Calculated from surface water concentrations, based on a wind speed of 5 m/sec.

+ Calculated based on habitat-specific fluxes and  regional habitat coverage data.

REFERENCES:
1:  Svensson and Rosswall, 1984             16:
2:  Svensson, 1976                         17:
3:  Sebacher et al., 1986                     18:
4:  Whalen and Reeburgh, 1988               19:
5:  Whalen and Reeburgh, 1992               20:
6:  Whalen and Reeburgh, 1990a             21:
7:  King, Quay, and Lansdown, 1989          22:
8:  Morrissey and Livingston 1992            23:
9:  Livingston and Morrissey,  in press          24:
10: Bartlett et al. 1992                      25:
11: Martens et al., in press                   26:
12: Fan et al. 1992                         27:
13: Ritter et al. 1992                        28:
14: Yarrington and Wynn-Williams, 1985       29:
15: Harriss et al., 1985
Crill etal., 1988
Verma et al., 1992
Dise, in press (a))
Vitt etal., 1990
Moore and Knowles, 1987
Moore, Roulet, and Knowles, 1990
Moore and Knowles, 1990
Roulet et al., 1992a
Clymo and Reddaway, 1971
Schiff etal., 1991
Ritter et al., 1991
Edwards etal., 1991
Hamilton et al., 1991
Roulet, pers. com.
                                                                          Page AP.14

-------
Appendix E: Methane Flux from Tussock Tundra (Warm Season Only)
             (Fluxes made using enclosure techniques
        unless otherwise noted; Flux units of mgCH4/m2/d)
MICROHABITAT
Ombrotrophic
communities
hummock
interhummock
shallow
depression
deeper depression
Summer tussock
tundra composite-
(1987)
Tussocks (1987)
Tussocks (1988)
Tussocks (1989)
Tussocks (1990)
Intertussocks (1987)
Intertussocks (1988)
Intertussocks (1989)
Intertussocks (1990)
Moss (1987)
Moss (1988)
Moss (1989)
Moss (1990)
Meadow/moist tundra
MP 298
MP346
MP416
MP416
MP298
MP318
MP346
WATER
LEVEL/INFO
471 1
12881
16911
25401
-
-
-10-92
-29-92
-12-112
-
-6-1 52
-29-1 02
0-202
-
-18-02
-40-42
-21-152
+ 2 cm
+ 2-5 cm
+ 3 cm
H20 sat.
-30 cm
-15 cm
-13 cm
AVG.
FLUX
1.1
3
16.5
25.8
22.4
29.5
60.8
48.8
96.8
2.8
25.2
12.4
5.9
0.9
10.1
29
3.8
32.6
77.6
152
53.7
3.1
0.3
12.5
STD.
DEV.
0.71
2.04
7.8
6.4
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
STD.
ERROR
0.32
1.02
3.5
2.9
-
-
-
-
-
-
-
-
-
-
-
-
-
-
14.1
8.5
5.4
3.1
0.3
-
REF.
1



2
3











4






                                                               Page AP-15

-------
MICROHABITAT
Moist tundra
Tussocks
Intertussocks
Upland tundra
(low tussocks)
Upland tundra
(low tussocks)
WATER
LEVEL/INFO
-
-
-
dry
dry
AVG.
FLUX
31
3.4
2.9
2.3
11
STD.
DEV.
19-523
-
-
-
-
STD.
ERROR
-
3.1
2.4
1.1
3
REF.
5
6

7
8
1 Avg. % water dry wgt.

2 Seasonal range in water table, all three habitat sub-sites.

3 95% confidence  interval.

REFERENCES:
1: Svensson and Rosswall, 1984
2: Whalen and Reeburgh, 1988
3: Whalen and Reeburgh, 1992
4: Sebacher et al., 1986
5: Whalen and Reeburgh, 1990
6: Morrissey and Livingston, in press
7: Bartlett et al.  1992
8: Fanetal. 1992
                                                                                            Page AP-16

-------
                   Appendix F:  Mean Regional Methane Fluxes (High Latitudes)
REGION SITE
Arctic Y-K1 Delta-
entire area
Y-K Delta-
flight paths
Y-K Delta-
tower area
Y-K Delta-
tower area
Y-K Delta-
tower area (I)2
tower area (2a)
tower area (2b)
North Slope
Boreal Schefferville
area
Schefferville
tower area
Schefferville
tower area
(July 17)
(July 25)
Schefferville
tower area
(July 17 &25)
Schefferville
flight paths
TECHNIQUE AVG. FLUX RANGE
(mgCH4/m2/d) (mgCH4/m2/d)
enclosures 48.5
aircraft eddy 44 25 - 85
correlation
enclosures 27
tower eddy 25
correlation
aircraft eddy
correlation - 40-60
59.7
72.1
enclosures 17.3
enclosures 1 8
tower eddy 1 6
correlation
tower eddy
correlation
19-24
10-29
aircraft eddy
correlation
24 - 29
aircraft eddy - 0-50
correlation
1Yukon-Kuskokwim Delta
2Numbers in parentheses indicate different flights and legs of flights.

Data from:
       Bartlett et al. 1992
       Fan etal. 1992
       Ritter et al. 1992
       Livingston and Morrissey, in press
       Moore, Roulet, and Knowles, 1990
       Edwards et al., 1991
       Ritter etal., 1991
       Schiff etal.,  1991
       Roulet, pers. com.
                                                                                         Page AP-17

-------
Appendix G: Rates of CH4 Uptake in Natural Ecosystems
    (Fluxes made using enclosure techniques unless
      otherwise noted; Flux units of mgCH4/m2/d)

            -- ARCTIC/SUBARCTIC WETLANDS --
HABITAT
Moist tundra meadow
Coastal tundra
Upland tundra
Alpine tundra
LATITUDE
53
-69
61
-68
AVG/MEDIAN RATE
-2.7
-0.86
-0.78
-0.2
RANGE
 '  ' - " " -- " - B" ' -
-0.5 --1.2
-0.2 --2.1
-
REFERENCE
1
2
3
4
             - ARCTIC/SUBARCTIC FORESTS -
HABITAT
Spruce forest
Floodplain taiga
Upland taiga
LATITUDE
-68
65
it
AVG/MEDIAN RATE
-0.3
-
-
RANGE
-
0--1.0
0--1.8
REFERENCE
4
5

                 - BOREAL WETLANDS -
HABITAT
Bog lagg
Rich fen
Forested swamp
Forested swamp
Domed bog
Fen
Bogs
Conifer swamps
Mixed hardwood
swamps
Thicket swamps
Marsh
Forests


LATITUDE
47
55
45

-
45
n

n
45
n
55
n
n
AVG/MEDIAN RATE
-
-
-
-
-
.
-
-
-
-
-
0.53
0.27
1.57
RANGE
-1.0
0--5
0--3
0--8
0--10
0- -0.2
0--0.1
0--0.2
0 - -5.8
0--0.3
0--0.3
-
-
-
REFERENCE
6
7



8





9


                                                           Page AP-18

-------
 -- TEMPERATE WETLANDS --
HABITAT
Ombrotrophic bog
Minerotrophic fen
Forested swamp
Salt marsh
Salt marsh
Salt marsh
Salt marsh
Salt marsh
LATITUDE
39

37
37
39
38
33
30
AVG/MEDIAN RATE
-
-
-
-1.4
-1.2
-1.8
-3.2
-3.2
RANGE
0 - -6.7
0- -7.7
-0.5 - -6
-0.7 - -2.7
-1 --1.4
-1.6 --2
-2.1 - -3.9
-
REFERENCE
10

11
12




  - TEMPERATE FORESTS -
HABITAT
Mixed forests
Red pine
Red spruce
Hardwood forest
Mixed forest
Mixed hardwood/
Conifer forest
Deciduous forest
Evergreen forest
Beech/spruce forest
Beech/spruce forest
Beech/oak/
maple forest
Spruce forest
Mixed forest
LATITUDE
39
-
-
43
43
440
42
H
49
-

-
"
AVG/MEDIAN RATE
-
-
-
-0.28
-1.6
-
-4.2
-3.5
-3.481
-3.451
-0.821
-0.251
-1.011
RANGE
0 - -6.5
0 - -6.5
0--1.2
-
0 - -4.9
-0.9 - -3.2
-3.5 --5.1
-3.2 - -4.2
-2.5 - -5.8
-1.9 - -5.8
-0.5 - -2.2
-0.1 --0.5
-0.5 --1.6
REFERENCE
10


13
14
15
16

17




-- TEMPERATE GRASSLANDS -
I HABITAT
Shortgrass steppe
LATITUDE
41
AVG/MEDIAN RATE
-0.61
RANGE
-0.35 - -0.84
REFERENCE |
18 |

                                            Page AP-19

-------
      -- TEMPERATE DESERT -
HABITAT
Sand w/sparse
vegetation
LATITUDE
37
AVG/MEDIAN RATE
-0.066
RANGE
0 - -4.38
1
REFERENCE
19
-- TROPICAL/SUBTROPICAL WETLANDS -
HABITAT
Open water
Non-forest swamp
LATITUDE
3
-
AVG/MEDIAN RATE
-
-
RANGE
-10.5
-11.3
REFERENCE
20
21
 - TROPICAL/SUBTROPICAL FORESTS -
HABITAT
=========
Dry flooded forest
Upland moist forest
Upland forest
Secondary forest
Tabunoco forest
Semi-deciduous
forest-oxisol soils
Semi-deciduous
forest-alfisol soils
Semi-deciduous
forest
Dry evergreen
forest:
Hillside
Valley
Virgin forest - clay
soils
Virgin forest - sandy
soils
LATITUDE
2
3
3
1
18
9
9
8
4
-
-
3
n
AVG/MEDIAN RATE
-1.9
-0.38
-0.71
-0.8
-0.58
-0.78
-0.35
-1.15

-1.7
-0.3
-0.78
-1.71
RANGE
^^sss^^ss^ss^sss^^^s^^ss^^^^
-0.8 - -4.6
-
-
-
-
-
-
-

-0.6 - -2.5
0--1.1
-0.36- -1.32
-1.57 --1.85
REFERENCE
!"" _ -  - '
22
13
23


24

25
26


27

                                                  Page AP-20

-------
                                 TROPICAL/SUBTROPICAL GRASSLANDS -
HABITAT
Dry savanna
Dry sandy savanna
Savanna - near
termite mounds
Broad-leaf savanna-
near termite
mounds
LATITUDE
0
5
39
25
AVG/MEDIAN RATE
-0.61
-0.07
-0.96
-1.25
RANGE
-0.46 - -0.81
-
-0.25 - -2.33
-0.31 - -2.45
REFERENCE
28

29
30
1  Calculated from soil gas profiles and radon exchange rates.

REFERENCES:
        1)   Whalen and Reeburgh, 1990a       17)
        2)   Kingetal., 1989                  18)
        3)   Bartlett et al.  1992                19)
        4)   Whalen and Reeburgh, 1990b       20)
        5)   Whalen et al., 1991                21)
        6)   Dise, in press (a)                  22)
        7)   Moore and Knowles, 1990         23)
        8)   Roulet etal.,  1992a               24)
        9)   Kingetal., 1991                  25)
        10)  Yavittetal., 1990                 26)
        11)  Harriss et al., 1982                27)
        12)  Bartlett et al., 1985                28)
        13)  Keller etal., 1983                 29)
        14)  Grill, 1991                        30)
        15)  King and Adamsen, in press
        16)  Steudler et al., 1990
Born etal., 1990
Mosier et al., 1991
Striegl etal., 1992
Bartlett et al., 1988
Bartlett et al., 1990
Tathy etal.,  1992
Keller etal.,  1986
Keller etal.,  1990
Scharffe etal., 1990
Delmas et al, 1992a
Goreau and de Mello, 1985
Delmas et al., 1991
Khalil etal., 1990
Seller etal.,  1984
                                                                                          Page AP-21

-------
s
O Q)
=;    co
o
co'
            ^ > rn c
            T> r^' n  i
      00
      c
      w
      13"
      CD
      CO
      c/>
   CD

   C
   CO
   CD
                  =R" CD
                  O  Q.
                     CO
      5- Q.

     % 33 _,  -
      O CO CD  P
      ^3 Q. 3  (D

     "n SI   w
      O ^. 
      O o no
         * ""*


      2^1
      |ol
      o^
rn
                 CQ
                  CD


                  O
         9 Tl
         CO 5
         en -^
                        oo
                     CD
                     co
                  >

                  m
                     -n
                     CD
                     CD
                     U)
                     CO
                     ci

-------