United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-88/048 Dec. 1988
&EPA Project Summary
Evaluation of Hydrologic
Models in the Design of
Stable Landfill Covers
Fairley J. Barnes and John C. Rodgers
Federal regulations stipulate that
landfill cover technology ensure the
long-term stability and integrity of a
hazardous waste landfill system.
Specific guidance for achieving
compliance with these regulations
requires cover designs that manage
the water balance on the landfill site.
Special attention must be given to
the design of the soil profile and the
establishment of a stable vegetative
cover in order to minimize erosion of
the surface soils and the percolation
of water into the waste. Recom-
mendations for a specific landfill
must be based on a combination of
field and laboratory data, and
computer modeling of water balance
to assess specific design scenarios.
The purpose of this study was to
investigate the utility of two hydro-
logic models (CREAMS and HELP) for
assisting in the process of designing
landfill cover systems that are stable
and free of maintenance require-
ments.
The process of parameterizing and
us'ng simple hydroloaic models is
outlined. Examples of modeling po-
tential and actual sites are pre-
sented. Results of the modeling
study suggest that, overall, CREAMS
performs more satisfactorily than
HELP in accurately predicting the soil
water storage in the soil profile
although more detailed calibration of
HELP will probably improve model
performance. While relative esti-
mates of runoff, deep percolation
and evapotranspiration are very
useful for comparing different cover
designs, absolute quantitative esti-
mates of these values are subject to
considerable error. Choice of values
for soil parameters requires more
experience and more detailed data
than is indicated by the docu-
mentation for either model. Data to
support choice of vegetation param-
eter values are largely unobtainable,
especially for the native plant
species which will inevitably invade a
landfill site. Successional processes
that lead to such invasions are
discussed, and it is concluded that
establishment of a vegetative com-
munity consisting of a mix of native
plant species would result in a
vegetative cover that is the most
stable and most efficient in removing
water from the soil profile.
This Project Summary was devel-
oped by EPA's Risk Reduction
Engineering Laboratory, Cincinnati,
OH, to announce key findings of the
research projects that are fully
documented in a separate report of
the same title (see Project Report
ordering information at back).
Introduction
Landfill systems for the long-term
containment of hazardous or municipal
waste will be an essential component of
hazardous waste management for the
foreseeable future. Even under regulatory
requirements for incinerating or de-
grading waste material in order to reduce
the toxicity of the waste, the remaining
sludges or residues must be disposed of
or stored in such a manner as to protect
the environment from toxic interactions.
-------
Certain hazardous wastes will probably
continue to be buried in landfills. In
addition, a large number of older waste
sites are in urgent need of closure
technology to prevent further damage to
the surrounding environment.
Major factors contributing to the mo-
bilization of toxic constituents in wastes
and their subsequent release to the
biosphere include the infiltration of
surface water, percolation of ground
water into the waste, and intrusion into
the waste by plants or animals. A critical
determinant of the long-term successful
application of land disposal technology is
the stability of the cap or cover of the
disposal unit. If, through a variety of
potential mechanisms such as erosion,
plant and animal disturbances, settling
and other mechanical disturbances, the
critical functions of the cover are com-
promised, then the long-term perform-
ance is compromised as well.
Establishment of a suitable and long-
lasting plant cover is the critical factor for
minimizing erosion of the surface soils of
the cover system and also for the
management of the water balance of the
entire cover system. The understanding
of the water balance components at a
site is essential for the design of a cover
system as well as for other remedial
measures. The soil which can act as
reservoir for water, or as a conduit for
subsurface transport of infiltrating water,
also supports the vegetative layer as a
mechanical base and is a reservoir for
water and nutrients. The water use and
rooting characteristics of the vegetation
can determine the extent of percolation
of water below the rooting zone as much
or more than soil hydrologic properties.
Thus, an estimate of the effects of
different soil and vegetation combina-
tions on the water balance of a cover
system must be an integral part of land
disposal design process.
The purpose of the project was to
investigate the utility of the hydrologic
models HELP (Hydrologic Evaluation of
Landfill Performance) and CREAMS (A
Field Scale Model for Chemicals, Runoff,
and Erosion from Agricultural Manage-
ment Systems) for assisting the process
of designing landfill cover systems that
are stable over long periods of time and
have no maintenance requirements in the
period after 30 years post-closure. The
report discusses concepts of low main-
tenance and stability for soil/vegetation
systems. The utility of these models was
explored through studies of the
parameters for the design of land
disposal units at several sites, assuming
a variety of regionally possible plant
covers, and in an assessment of the
performance of the models in simulating
experimental landfill cover systems
under a variety of controlled field condi-
tions of vegetative cover.
Vegetative Succession on
Landfill Sites
Although it has been recognized that
the vegetative cover is an important
component in cover design, the ques-
tions of long-term stability of the
vegetation community on the landfill (and
hence its long-term performance) is
rarely addressed. Recommendations
have been made to use cultivated or
domesticated species, particularly
grasses, for reclamation purposes. This
is largely because many grass species
tend to rapidly form a sod which can
effectively protect the soil surface from
erosion, and also because they have
shallow rooting patterns and so are less
likely to invade the waste than deeper
rooted species. Although a grass cover
may be an attractive alternative for the
short-term, successional processes will
inevitably lead to invasion of the site by
native species from the surrounding area.
A few current documents suggest the
use of native species where appropriate
but successional processes per se, and
their effect on cover integrity and
effectiveness, have not been addressed.
Few data are available on important
native species which might be candi-
dates for vegetative cover treatments.
The specific course of succession can
only be predicted from studies of specific
habitats. In particular, studies on
drastically disturbed areas would be
invaluable. Without management inter-
vention, the soil environment of a landfill
cover is not likely to resemble the soil
profile of a natural or mildly disturbed
community for many years. Not only are
the physical characteristics of the soil
layers (porosity, bulk density, hydraulic
conductivity) changed by the construc-
tion process, but the organic content is
low until the soil biota reaches levels
comparable to those in natural com-
munities.
The plant cover on a landfill site will
inevitably undergo successional
changes. In the short term, relatively
rapid changes in species and population
sizes will result in a relatively unstable
vegetative community. In the long term,
dominant species from surrounding
natural vegetation will invade the site to
the extent that the soil environment
resembles that of a natural community in
terms of depth, layering, texture, water
holding capacity, nutrient status, and s
organisms. Multiple pathways of si
cession are possible, indeed likely,
any given site. Management interventi
can result in a vegetative commun
being established that is most advan
geous for long-term, low-maintenan
equilibrium. Such management de
sions, however, must be based on
realistic assessment of the life histc
characteristics and physiological toll
ances of the species under consideratic
Competitive interaction between speci
should also be considered in choosi
species mixes. Such data are oft
unavailable for native species. A study
the plant associations present in
particular locale, along with a history
land usage, can often supply
reasonable estimate of the possit
stages of succession, and perhaps ev
of the time course of the process. Me
detailed assessment of the soil conditic
and microclimatic environment of
desired vegetative phase should I
undertaken. These data will provide
starting point for designing establishme
procedures necessary to accelerate
decelerate successional processes so
to maximize the time the vegetative cov
has an optimum mix of species.
Simulation of Potential Landfill
Sites
Understanding and controlling wat
balance relationships in hazardous was
landfill cover treatments is an essent
component of the design and operatii
of hazardous waste management sites.
many climatic zones evapotranspiratk
accounts for the largest output cor
ponent of the water balance equation,
AS = P-Q-ET-L
where over a time interval At, AS
change in soil moisture content, P
precipitation, Q = runoff, ET = evap
transpiration, and L = seepage. It
particularly important to understand U
interrelationships between meteorolog
vegetation, and soils that determir
evapotranspiration so that the seepac
component be estimated with confidenc
The development of computer simulatic
models of the hydrologic cycle offers tt
potential for simulating the several critic
processes involved in tracking hydrolog
responses to daily (or smaller) time stef
of water input. In the present study tr
USDA model CREAMS and the EP
model HELP were used to modi
hydrologic processes over varying tirr
periods. These are physically base
-------
models and so should not require
calibration for each application. Field
scale hydrologic responses are based on
models for infiltration, soil water move-
ment, and soil/plant evapotranspiration
between storms. Processes are con-
tinuously simulated on a daily time step
for evapotranspiration and soil water
movement between storms. The simu-
lation of water redistribution between
storms provides the basis for prediction
of seepage below the root zone.
The CREAMS model was intended for
modeling field-scale agricultural sys-
tems. The model has been used in
several areas of waste management
research in semi-arid climates, includ-
ing erosion studies, water balance and
primary production of desert shrubs and
landfill cover design. In this study it was
used to evaluate two potential landfill
sites, one in Houston, Texas and the
other in Fresno, California. Both are
areas with a history of problems with
hazardous waste management and are
on the Hazardous Waste Site National
Priority List. Together with the Los
Alamos experimental site, a wide
diversity of climatic regimes, soil types,
and native vegetation are represented by
the modeling studies.
CREAMS was used in the daily
rainfall-runoff mode to obtain a com-
plete water budget (estimates of runoff,
evapotranspiration, percolation and soil-
water storage, or water content in the soil
column to the depth of the rooting zone)
on each day that there was a pre-
cipitation event. Monthly and annual
water budgets are also obtained. The
model is one-dimensional, treating only
the process of vertical transport of water
in the soil column. Lateral inflow or
drainage are not accounted for.
The alternative model, HELP, devel-
oped by EPA, is directly applicable to
most landfill designs. In contrast to the
purely one-dimensional character of
CREAMS, the HELP model permits
quasi-two-dimensional modeling of
soil water routing by including lateral flow
simulation in drainage layers. Precip-
itation inputs are modeled one-dimen-
sionally down to the depth of lateral
drainage and impermeable membrane
liners. Lateral flow out of drainage layers
is treated separately. The infiltration
routine is similar to that used in
CREAMS, and there are changes
(claimed to be improvements) in the
treatment of percolation and evapo-
transpiration. The model is interactive
and user-friendly, with default climatic
and soils data available for many regions
in the United States Default estimated
vegetation data are also available to the
user. Alternatively, the user can specify
exact parameter values specific to the
site being modeled. Both models require
soil characteristics, seasonal vegetation
characteristics and data, and soil cover
design data as input. Monthly mean
temperatures, mean monthly solar radi-
ation values, and daily precipitation
inputs are required.
A general plan for the evaluation of a
potential landfill site and for acquiring the
data for a modeling exercise is as fol-
lows.
1. Background information, including
history and setting, climate, vegetation
and soils, should be collected. This
step is particularly important to
provide a basic set of resources for
the following data requirements and
also to supply perspective to the
entire exercise.
2. Examine climatic data, including 20 to
30 year records of temperature, pre-
cipitation and solar radiation. Such
records are available from local or
regional weather bureau stations or
the U.S. Department of Commerce.
3. Examine soil data from the relevant
USDA Soil Survey. Local sources of
top soils, barrier soils (clays), fill soils
(subsoils), and "special-needs" soils
(cobbles, gravels, sands) must be
located and the textural classes of the
soils determined. Extra effort should
be expended to determine the actual
hydrologic properties of the soils:
hydraulic conductivities, field capacity,
wilting point and curve number under
different topographic and vegetative
conditions.
4. Evaluate the vegetation of the region.
The results of this study strongly
suggest that vegetation of the landfill
be accomplished with native species
which are preadapted to the soils and
climate of the region and are thus
most likely to form a stable vegetative
community for as long as possible.
The species present in the different
vegetative communities should be
determined, along with the relative
importance of the species Likely
successional sequences should be
evaluated, as well as the approximate
timing of the various stages under
natural conditions, and if at all pos-
sible, an estimate of how succession
is modified by management practices
Details on leaf area indices, above
ground biomass, seasonal variabilities,
rooting structures, and physiological
characteristics (especially transpiration
rates) should be compiled for the
dominant members of each com-
munity type. Lists of species present
in the various communities are rela-
tively easy to obtain from floras of the
state or region. However, all other
attributes are generally very difficult to
obtain, and are often unavailable. If the
data is not forthcoming from these
sources, limited field studies were
indicated.
5. The selection of a particular model
must depend on demonstrating the
appropriateness 'of the model to
addressing the objectives. In the
cases discussed in this report, the
objectives were to specifically evalu-
ate the effects of varying soils and
vegetation characteristics on water
balance of the landfill cover system,
and thus the model results had to
supply estimates of percolation,
evapotranspiration, runoff, and soil
water storage. In addition, the model
could not be so "data hungry" that
input data was impossible to obtain
These considerations resulted in
choosing one-dimensional hydrologic
models with a high degree of flexibility
in changing input parameters as well
as a "user-friendly" operating envi-
ronment.
6. If at all possible the model should be
calibrated using data from a com-
parable landfill site or natural area in
the same region.
7. The time scale of the model simu-
lations must be considered with
reference to the desired objectives. In
an exercise to determine possible
worst case scenarios in the event of
unusually severe climatic conditions,
the meteorological data base should
be sufficiently long so as to incor-
porate such severe conditions
8. The interpretation of the results must
be relevant to both the objectives of
the study and the capabilities of the
model. It may be unrealistic to expect
absolute quantitative accuracy from
model predictions. However, the
values of the results can indicate with
quite good accuracy the relative utility
of design scenarios.
9. A major issue in a modeling exercise
is to take the modeling results, along
with other information and then design
an actual vegetative cover of specific
ratios of plant species. This may be
the most difficult step in the evaluation
of potential landfill design. The
-------
success of this step depends largely
on the accuracy and depth of the
information gathered during steps 3
and 4 above.
The results of the modeling studies on
two potential landfill sites showed that
generally increasing cover thickness
alone is not an efficient way of reducing
deep percolation. If a poor selection of
vegetative treatment is made, large
amounts of water may be available for
percolation at certain times of the year,
and the simple expedient of increasing
cover thickness is not likely to be an
efficient remedy, particularly by com-
parison with use of low permeability or
impermeable barriers. However, in-
creasing the biomass of vegetation on a
site will significantly increase ET and
decrease percolation on a yearly basis.
Seasonally, the form and phenology
(seasonal activity) of the vegetation has a
very significant effect on the size and
frequency of percolation events. These
results suggest that it is most important
to select a variety of species to
revegetate a site, so that different forms
(trees, shrubs, forbs and grasses) and
different phenologies (evergreen, decid-
uous, warm-season, cool-season) are
well-represented. The ratio of these
factors would ideally reflect the seasonal
distribution of precipitation events.
Mitigation of erosion will depend on
having a vigorous plant cover throughout
the year. The size of runoff events as
modeled by CREAMS was not greatly
affected by the biomass of the vegeta-
tion.
The advantages of computer model
simulation of cover system performance
are many. Once site data are assembled
a large number of "what if" questions
can be explored systematically, and the
outcomes can be compared quanti-
tatively. However, data deficiencies can
be a serious problem leading to
predicting results of uncertain value. Also
a number of critical biological and geo-
chemical questions about cover design
and performance may be raised but
cannot be readily answered by these
models.
Simulation of Experimental
Landfill Covers
Field Experiment
A study was undertaken on a low-
level radioactive waste landfill site ("Area
B") in Los Alamos, New Mexico. About
half of the site had in place a layered soil
cover consisting of local topsoils and
crushed tuff. This type of soil profile has
been the standard cover system on
landfill sites at Los Alamos, and therefore
a significant amount of data were already
available on its physical and ecological
characteristics. This profile was in two
areas at Area B, termed "east control"
and "west control". About 25% of the
area of Area B had a soil profile
consisting of 15 cm top soil, 50 cm
crushed tuff, and a gravel-cobble layer
designed to act as a biobarrier to prevent
animal and root intrusion into the waste
fill. This type of profile (termed "central
biobarrier") has been studied at Los
Alamos, and a considerable amount of
information about its performance has
been gained through erosion and
subsurface studies. Furthermore, exactly
such a layer combination has been
recommended to EPA as a means for
providing a gas channel for controlled
release under tight soil caps. The profiles
had been in place for a year before the
start of the EPA-sponsored study.
East and west control profiles differ in
the amount of sandy clay loam in the
profile. The west profile is more typical of
landfill covers at Los Alamos, having
about 15 cm of top soil over 85 cm of
crushed tuff. The east profile has a much
higher amount of top soil (sandy clay
loam) mixed into the profile as a result of
the reconstruction in 1982, and thus the
east soil profiles have significantly higher
water holding capacity than the west soil
profile.
Study plots were established on each
profile with bare, grass and shrub
vegetative covers. The dense rabbit-
brush planting had shrubs planted on
0.68 m (2.2 ft) centers (2.15 plant/m2),
and the sparse rabbitbrush planting at
1/5 density had shrubs planted on 1.5 m
(5 ft) centers (0.43 plants/m2).
In each plot, three or four neutron
access tubes were installed and soil
moisture distribution with depth meas-
ured with a neutron moisture probe
throughout the year.
Determining absolute values and
seasonal changes in biomass and leaf
area index (LAI) on each study plot was
an important subtask in this project. LAI
is the total projected leaf area (one side
only) per unit area of ground. The
CREAMS and HELP models require an
estimate of LAI as input for calculating
utilization of soil moisture by the
vegetative cover, and these values have
large seasonal variations. Hence, it was
necessary to develop nondestructive
methods that allow frequent estimates
during the year without disturbing the
vegetative cover.
For the shrubs and grasses, d
structive sampling was used to determii
the dry leaf mass as a fraction of c
mass of current growth and the spec!
leaf mass (Table 1). Regression mod<
were developed relating crown bioma
to crown volume for the shrubs ai
canopy biomass to percent ground cov
for the grasses (Table 2). Using tl
relationships developed, total bioma
and leaf area indices were estimati
during the growing season usii
nondestructive measurements of shn
volume or estimates of grass canoi
ground cover.
Field Results
Both soil profile composition ai
vegetative cover had a strong mfluen<
on soil water storage in the rooting zoi
(0 to 70 cm). Of the three soil profile
the east control profile (high clay contei
retained more moisture than the we
control profile (low clay content) or tl
central biobarrier profile (shallow soil ov
a gravel-cobble capillary barrier laye
(e.g., Figure 1). Soil moisture in tl
central profile decreased to very Ic
levels during each growing season. Tr
is most likely the result of efficient mmir
of soil water by the shrub roots in tl
restricted soil profile as well as latei
drainage of soil water in the soil abo1
the gravel-cobble layer which served ,
a barrier to capillary moisture flow.
Volumetric soil water was lowest und
vegetative cover of shrubs (e.g., Figu
2). After 2 years growth, bare plots had
to 7 cm more water in the assume
rooting zone (70 cm) than the shn
plots. Grass plots were intermediate
soil moisture. The maximum rate
decline in soil moisture was usual
observed on the shrub plots, with gras
plots having an intermediate rate ar
bare plots lowest. The shrubs are knov
to be quite deeply rooted. The effect
this was apparent m that soil moisture
60 to 80 cm depth decreased under tr
shrub cover during the summer month
In contrast, water withdrawal by the gra:
cover was not apparent at these depths.
Simulation of Field Experiment
Simulations of the different scenanc
were made using the CREAMS ar
HELP models. Parameter values for sc
hydrologic characteristics were optimize
using data for a shrub plot for whic
there was a four-year record of sc
moisture (two years prior to the start
the experiment plus two years during th
study). Optimization was performed t
-------
Table 1. Biomass and Leaf Characteristics for Shrubs and Mixed Grasses
Rabbitbrush
specific leaf mass (g/cm2)
leaf mass/total biomass
Mixed grass
specific leaf mass (glcm2)
leaf mass/total biomass
Mean
Value
0.0169
0.499
0.0243
0.424
Standard
Error
0.0013
0.021
0.0021
0.023
Number of
Samples
33
33
122
10
Table 2. Regression Relationships for Shrub and Grass Cover at Area B
Cover Equation Y X
Shrubs y = 43.43 + 1.57x biomass (g)
Grass y = 1.28 + 0.55x biomass (g/m2) % cover
shrub volume
(m3 x 103)
0.98
0.66
Area B Shrub Plots
35
I
30 1
25 -
20 -
15 -
10 -
J F M A M J J A SONDJFMAMJJ A S 0 N
1985 1986
Figure 1. Volumetric soil moisture (average of 20, 40 and 60 cm depths) on shrub plots
at Area fi over 2 years.
-------
Area B Bast Control Plots
35
30-
25 -
20 -
15 -
10
A Bare
Q Grass
O Shrub
.I...I...I...I...I...I...I...I...I...I...I...I...I...I...I...I.
JFMAMJJASO NDJFMAMJJASON
1985 1986
Figure 2. Volumetric soil moisture (average of 20, 40 and 60 cm depths) on plots on the
east soil profile at Area B.
varying values within recommended
ranges for hydraulic conductivity, field
capacity, wilting point and curve number
and then performing a linear regression
analysis between observed and
predicted soil moisture over a three-
year period. The process was repeated
until it was apparent which combination
of parameter values resulted in the
highest correlation coefficient (indicating
the least amount of scatter in values), the
slope of the regression line closest to 1.0
(indicating a dynamic range in predicted
values that most closely approximates
the variability observed in the field) and
an intercept closest to 0.0 (indicating an
absolute value of soil moisture values
most closely resembling the observed
values).
Simulation of all treatment scenarios
were performed for a two-year period
with CREAMS (using optimized soil
parameters), and with HELP (using
optimized soil parameter values and also
again using the default parameter values
as supplied by the HELP model). The
overall predictive power of each model
was assessed by a linear regression
analysis of predicted versus observed
soil moisture over all plots (summarized
in Table 3).
Predications of soil moisture using the
CREAMS model more closely resembled
the measured soil moisture over a range
of soil and vegetation treatments than
predictions obtained using the HELP
model using either set of soil parameter
values. The HELP model produced more
accurate predicted values for soil
moisture using the default soil parameter
values.
Simulations of the soil profile treat-
ments which more closely resembled a
natural profile were more accurate than
those of the profiles which had a more
artificial layered structure. Treatments
with shrub or bare covers were more
accurately modeled than treatments with
grass covers.
The optimization process of CREAMS
narrowed the range of choices of several
parameters, and the final selection of
parameter values departed from the
average values in a meaningful way. The
most successful HELP runs were cor
ducted with the default options availabl
in the program itself, resulting in
relatively naive selection of paramete
values and low correlation betwee
predicted and observed soil moisture.
is noteworthy that CREAMS runs cor
ducted with average parameter value
based on soil texture classes als
resulted in poor correlation coefficient
for the comparison between observe
and predicted soil moisture values. It i
quite likely that calibrating the HEL
model with particular reference to laten
transport, seepage and runoff (obsei
vations which were not available from th
field experiments in this study) woul
substantially improve the performance (
the model.
Although CREAMS more accurate!
predicted field observations than th
HELP model, CREAMS still should b
used with several precautions. Actu;
values for soil hydrologic parameters wi
undoubtedly produce better results tha
values assumed on the basis of so
texture class. However, actual values fc
-------
the soils to be used on the landfill will, in
all probability, not be readily available. A
reasonable effort to develop values for
soil parameters through laboratory tests
would greatly increase the accuracy of
the model.
Modifying vegetation parameters and
soil layer depths and then assessing the
resultant model predictions for perco-
lation events should indicate the mini-
mum levels of plant LAI necessary to
control percolation at different times of
the year and the appropriate soil layer
depths. Implementing the soil profile
design is trivial compared to choosing
species that would rapidly produce the
desired LAI values within the seasonal
constraints suggested by the modeling
exercise. Site specific field studies on
natural communities in the area of the
proposed landfill site are needed to
determine what species make up
different communities, the possible suc-
cessional communities, and the natural
timing of succession. Some of this in-
formation could be obtained from
literature or anecdotal research, while
field studies might be necessary for the
rest.
Conclusions
The results of this study have
demonstrated unequivocally that the role
of native vegetation in determining site
stability and integrity must be evaluated
and considered in designing a hazardous
waste landfill cover system. It is apparent
that native species are much hardier than
cultivated species. Given a long enough
period of time with no human inter-
ference (i.e. no maintenance) native
species will invade and colonize a landfill
site, and some sequence of succession
will proceed thereafter. Unfortunately,
these processes are not well docu-
mented, especially for severely disturbed
sites or on constructed soil profiles.
The use of simple hydrologic models
to simulate the water balance of landfill
cover systems can assist the process of
designing soil and vegetative systems for
site closure. Two models (CREAMS and
HELP) were used to simulate eight
experimental cover designs on a landfill
site at Los Alamos, New Mexico. Results
of the modeling study suggest that,
overall, CREAMS performs more satis-
factorily than HELP in accurately pre-
dicting soil water storage in the soil
profile, although more detailed calibration
of HELP will probably improve model
performance. Careful parameterization of
the models was a key to successful
simulation. Few data exist for leaf area
indices and rooting depths of native plant
species. Values of hydrologic parameters
of soils may be quite different in the
constructed soil profiles of landfill covers
than under laboratory or natural condi-
tions. These areas are in critical need of
further research.
Table 3. Linear Regression Relationships (y = a * bx) Between Observed Average Soil Moisture from 0-70 cm (x) and Soil Moisture (y) as
Predicted by CREAMS (Using Optimized Parameters) and HELP (Using Default Parameters)
Plot
2
3
4
7
10
11
12
13
All Plots
East Plots4
West Plots5
Shrub Plots6
Grass Plots7
Bare Plots^
Treatment
Bare
Shrub1
Grass
Shrub-Cobble2
Sparse Shrub3
Shrub
Grass
Bare
a
13.04
-3.04
11.90
5.1S
13.87
-0.89
20.49
22.07
3.83
-1.73
9.34
4.14
20.52
8.94
CREAMS Model
b
0.59
1.09
0.54
0.68
0.46
0.87
0.27
0.10
0.82
1.06
0.58
0.75
0.24
0.69
r2
0.61
0.62
0.72
0.50
0.72
0.60
0.68
0.14
0.62
0.71
0.29
0.56
0.45
0.56
a
5.93
11.09
11.36
13.77
17.87
15.66
15.86
10.75
12.47
13.36
16.85
12.40
14.57
-1.39
HELP Model
b
0.58
0.47
0.40
0.17
0.14
0.23
0.22
0.24
0.33
0.34
0.14
0.35
0.28
0.79
r*
0.42
0.63
0.25
0.33
O.T3
0.26
0.26
0.26
0.39
0.36
0.05
0.48
0.31
0.58
1 Shrub density 2.15 plantslm2
2Shrub density 2.15 plantslm2 over a soil profile with a cobble-gravel layer
3Shrub density 0.43 plantslm2
"Plots 2,3,4
5Plots 10,11,12,13
6Plots3,10, 11
>7Plots 4,12
BPIotS 2,13
-------
Fairley J. Barnes and John C. Rodgers are with the Los Alamos National
Laboratory, Los Alamos, NM 87545.
Naomi P. Barkley is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of Hydrologic Models in the Design of
Stable Landfill Covers," (Order No. PB 88-243 811/AS; Cost: $21.95, subject to
change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-88/048
0001961 HWEft 5N
N V
"OS DEARBORN ST
CHICAGO
fu
60604
------- |