Attribute_Accuracy_Report:
Samples were obtained from the dust traps by carefully washing
the marbles, screen, and pan with distilled water into plastic
liter bottles. In the laboratory, the sample was gradually
dried at about 35°C in large evaporating dishes; coarse
organic material is removed during this process. Subsequent
analyses on dust samples included, in the order they were
performed: (1) moisture, (2) organic matter, (3) soluble
salts and gypsum, (4) total carbonate (calcite plus
dolomite), (5) grain size, (6) major-oxide chemistry, and
(7) mineralogy (sand, silt, and clay fractions). The database
for any given site commonly contains gaps depending on how far
the sample for a particular year could be stretched through
the analytical cascade. In some cases, samples from different
years at the same site or adjacent sites were combined to
obtain enough material for measuring grain size.
A sample was commonly retrieved and used in more than one
analysis if the first analytical procedure used was non-
destructive. These sequential analytical techniques
included: (1) Moisture and organic-matter content (Walkley-
Black procedure in Black, 1965) were measured on the same
split using 0.05 g. (2) The entire sample was used to extract
the solution to measure soluble salts (Jackson, 1958) and was
then dried and recovered; thus, subsequent analyses were
performed on samples without soluble salts. (3) A 0.25-g
split was used to analyze total carbonate (Chittick procedure
in Singer and Janitzky, 1986). This split, free of carbonate
after the analysis, was recovered and used to analyze for
major oxides and zirconium. (4) When sufficient sample (0.4g)
existed to obtain grain size using the Sedigraph rather
than by pipette analysis, the clay and silt fractions were
saved and used to analyze mineralogy by X-ray diffraction.
Most of the laboratory analyses were performed in the
Sedimentation Laboratory of the Institute of Arctic and Alpine
Research in Boulder, Colorado, using standard laboratory
techniques for soil samples (see Black, 1965, and Singer and
Janitzky, 1986) that we adapted for use on very small samples
(the non-organic content of a dust sample collected from one
trap typically weighs less than 1 g/yr). These adaptations
generally result in larger standard errors than normal for the
results of different techniques because the amount of sample
used is smaller than the recommended amount.
The sampling design for this study was not statistically based;
rather, sites were chosen to provide data on dust influx at soil-
study sites and to answer specific questions about the relations
of dust to local source lithology and type, distance from source,
and climate. Some sites were chosen for their proximity to
potential dust sources of different lithologic composition (for
example, playas versus granitic, calcic, or mafic alluvial fans).
Other sites were placed along transects crossing topographic
barriers downwind from a dust source. These transects include
sites east of Tonopah (43-46) crossing the rhyolitic Kawich Range,
sites downwind of northern (40, 35, 36) and central Death Valley (
38, 39, 11-14) crossing the mixed-lithology Grapevine and Funeral
Mountains, respectively, and sites downwind of Desert Dry Lake
crossing the calcareous Sheep Range (47-50) north of Las Vegas.
In addition, some sites were chosen for their proximity to weather
stations.
Specific locations for dust traps were chosen on the basis of the
above criteria plus accessibility, absence of dirt roads or other
artificially disturbed areas upwind, and inconspicuousness. The
last factor is important because the sites are not protected or
monitored; hence, most sites are at least 0.5 mile from a road or
trail. Despite these precautions, dust traps are sometimes
tampered with, often violently. This is a particular problem in
areas close to population centers, and most of these sites (52-55
near Los Angeles and 17-19 and 22 near Las Vegas) have been
abandoned. A few other sites, mostly those that appeared to be
greatly influenced by nearby farming (20, 21, and 41), were
eliminated in 1989. Dust traps were also generally placed in
flat, relatively open areas to mitigate wind-eddy effects created
by tall vegetation or topographic irregularities.
See notes in the Attribute_Accuracy_Report regarding combination
of samples too small for individual analyses. Generally the data
from ICP, oxides, and mineralogy are for combined samples.
The 55 sites established in 1984 and 1985 were sampled annually
through 1989 in order to establish an adequate statistical basis
to calculate annual dust flux. Sampling continues at 37 of these
sites (many sites now have two or more dust traps) every two or
three years as opportunity and funding permit.
The most important factors that influenced dust-trap design in
this study were: (1) measuring the amount of dust added to
soils; (2) sampling on an annual basis; (3) no protection other
than being hard to find; and (4) the cost and ready availability
of components that might have to be replaced from sources in small
towns. The original design consists of a single-piece Teflon-
coated angel-food cake pan (see note 1) painted flat black on the
outside to maximize water evaporation and mounted on a steel fence
post about 2 m above the ground. A circular piece of 1/4-inch-
mesh galvanized hardware cloth is fitted into the pan so that it
rests 3-4 cm below the rim, and glass marbles fill the upper part
of the pan above the hardware cloth. The Teflon coating is non-
reactive and adds no mineral contamination to the dust sample
should it flake. The hardware cloth resists weathering under
normal conditions. The 2-m height eliminates most sand-sized
particles that travel by saltation rather than by suspension in
air; sand grains are not generally pertinent to soil genesis
because they are too large to be translocated downward into soil
profiles. The marbles imitate the effect of a gravelly fan
surface and prevent dust that has filtered or washed into the
bottom of the pan from being blown away. The empty space below
the hardware cloth provides a reservoir that prevents water from
overflowing the pan during large storms. This basic design was
modified in 1986 in two ways. In many areas, the traps became
favored perching sites for a wide variety of birds. As a result,
significant amounts of non-eolian sediment were locally added to
the samples (as much as five times the normal amount of dust at
some sites). All dust traps were fitted with two metal straps
looped in an inverted basket shape over the top and the top
surfaces of the straps were coated with Tanglefoot1. This sticky
material never dries (although it eventually becomes saturated
with dust and must be reapplied) and effectively discourages birds
from roosting. In addition, extra dust traps surrounded by alter-
type wind baffles were constructed at four sites characterized by
different plant communities. These communities and sites are:
blackbrush (Coleogyne ramosissima), creosote bush (Larrea
divaricata), and other low brushy plants at sites 1-5 on Fortymile
Wash; Joshua tree (Yucca brevifolia), other tall yucca species,
and blackbrush at site 18 on the Kyle Canyon fan; pinyon-juniper
(Pinus monophylla-Juniperus sp) at site 7 on Pahute Mesa; and
acacia (acacia sp), creosote bush, and blackbrush at site 26 near
the McCoy Mountains. The wind baffles imitate the effect of
ground-level wind speed at the 2-m height of the dust trap and
permit comparison of the amount of dust caught by an unshielded
trap with the amount that should be caught at ground level where
vegetation breaks the wind.
Source_Information:
Source_Citation:
Citation_Information:
Originator: National Climatic Data Center
Publication_Date: 1992
Title: California: Climatological Data Annual Summary
Type_of_Source_Media: paper
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1961
Ending_Date: 1990
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: NCDC 61-90 A CA
Source_Contribution:
Data used to calculate mean annual temperature (MAT) and
mean annual precipitation (MAP) at dust trap sites
Source_Information:
Source_Citation:
Citation_Information:
Originator: National Climatic Data Center
Publication_Date: 1992
Title: Nevada: Climatological Data Annual Summary
Type_of_Source_Media: paper
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1961
Ending_Date: 1990
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: NCDC 61-90 A NV
Source_Contribution:
Data used to calculate mean annual temperature (MAT) and
mean annual precipitation (MAP) at dust trap sites
Source_Information:
Source_Citation:
Citation_Information:
Originator: National Climatic Data Center
Publication_Date: 1992
Title:
California: Monthly station normals of
temperature, precipitation, and heating and
cooling degree days 1961-1990
Series_Information:
Series_Name: Climatography of the United States
Issue_Identification: 81
Type_of_Source_Media: paper
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1961
Ending_Date: 1990
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: NCDC 61-90 M CA
Source_Contribution:
Data used to calculate mean annual temperature (MAT) and
mean annual precipitation (MAP) at dust trap sites
Source_Information:
Source_Citation:
Citation_Information:
Originator: National Climatic Data Center
Publication_Date: 1992
Title:
Nevada: Monthly station normals of temperature,
precipitation, and heating and cooling degree days
1961-1990
Series_Information:
Series_Name: Climatography of the United States
Issue_Identification: 81
Type_of_Source_Media: paper
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1961
Ending_Date: 1990
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: NCDC 61-90 M NV
Source_Contribution:
Data used to calculate mean annual temperature (MAT) and
mean annual precipitation (MAP) at dust trap sites
Source_Information:
Source_Citation:
Citation_Information:
Originator: U.S. Department of Commerce (Weather Bureau)
Publication_Date: 1964
Title:
California: Climatic summary of the United States--
Supplement for 1951 through 1960
Series_Information:
Series_Name: Climatography of the United States
Issue_Identification: 86-4
Type_of_Source_Media: paper
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1951
Ending_Date: 1960
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: DOC 51-60 CA
Source_Contribution:
Data used to calculate mean annual temperature (MAT) and
mean annual precipitation (MAP) at dust trap sites
Source_Information:
Source_Citation:
Citation_Information:
Originator: U.S. Department of Commerce (Weather Bureau)
Publication_Date: 1964
Title:
Nevada: Climatic summary of the United States--
Supplement for 1951 through 1960
Series_Information:
Series_Name: Climatography of the United States
Issue_Identification: 86-4
Type_of_Source_Media: paper
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1951
Ending_Date: 1960
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: DOC 51-60 NV
Source_Contribution:
Data used to calculate mean annual temperature (MAT) and
mean annual precipitation (MAP) at dust trap sites
Process_Step:
Process_Description:
The most important factors that influenced dust-trap
design in this study were: (1) measuring the amount of
dust added to soils; (2) sampling on an annual basis; (3)
no protection other than being hard to find; and (4) the
cost and ready availability of components that might have
to be replaced from sources in small towns. The original
design consists of a single-piece Teflon-coated angel-food
cake pan (see note 1) painted flat black on the outside to
maximize water evaporation and mounted on a steel fence
post about 2 m above the ground. A circular piece of 1/4-
inch-mesh galvanized hardware cloth is fitted into the pan
so that it rests 3-4 cm below the rim, and glass marbles
fill the upper part of the pan above the hardware cloth.
The Teflon coating is non-reactive and adds no mineral
contamination to the dust sample should it flake. The
hardware cloth resists weathering under normal
conditions. The 2-m height eliminates most sand-sized
particles that travel by saltation rather than by
suspension in air; sand grains are not generally pertinent
to soil genesis because they are too large to be
translocated downward into soil profiles. The marbles
imitate the effect of a gravelly fan surface and prevent
dust that has filtered or washed into the bottom of the
pan from being blown away. The empty space below the
hardware cloth provides a reservoir that prevents water
from overflowing the pan during large storms. This basic
design was modified in 1986 in two ways. In many areas,
the traps became favored perching sites for a wide variety
of birds. As a result, significant amounts of non-eolian
sediment were locally added to the samples (as much as
five times the normal amount of dust at some sites). All
dust traps were fitted with two metal straps looped in an
inverted basket shape over the top and the top surfaces of
the straps were coated with Tanglefoot. [Use of trade
names by the U.S. Geological Survey does not constitute an
endorsement of the product.] This sticky material never
dries (although it eventually becomes saturated with dust
and must be reapplied) and effectively discourages birds
from roosting. In addition, extra dust traps surrounded
by alter-type wind baffles were constructed at four sites
characterized by different plant communities. These
communities and sites are: blackbrush (Coleogyne
ramosissima), creosote bush (Larrea divaricata), and other
low brushy plants at sites 1-5 on Fortymile Wash; Joshua
tree (Yucca brevifolia), other tall yucca species, and
blackbrush at site 18 on the Kyle Canyon fan; pinyon-
juniper (Pinus monophylla-Juniperus sp) at site 7 on
Pahute Mesa; and acacia (acacia sp), creosote bush, and
blackbrush at site 26 near the McCoy Mountains. The wind
baffles imitate the effect of ground-level wind speed at
the 2-m height of the dust trap and permit comparison of
the amount of dust caught by an unshielded trap with the
amount that should be caught at ground level where
vegetation breaks the wind.
Process_Date: 1984
Process_Step:
Process_Description:
Samples were obtained from the dust traps by carefully
washing the marbles, screen, and pan with distilled water
into plastic liter bottles. In the laboratory, the sample
was gradually dried at about 35°C in large evaporating
dishes; coarse organic material is removed during this
process. Subsequent analyses on dust samples included, in
the order they were performed: (1) moisture, (2) organic
matter, (3) soluble salts and gypsum, (4) total carbonate
(calcite plus dolomite), (5) grain size, (6) major-oxide
chemistry, and (7) mineralogy (sand, silt, and clay
fractions). The database for any given site commonly
contains gaps depending on how far the sample for a
particular year could be stretched through the analytical
cascade. In some cases, samples from different years at
the same site or adjacent sites were combined to obtain
enough material for measuring grain size.
A sample was commonly retrieved and used in more than one
analysis if the first analytical procedure used was non-
destructive. These sequential analytical techniques
included: (1) Moisture and organic-matter content (Walkley-
Black procedure in Black, 1965) were measured on the same
split using 0.05 g. (2) The entire sample was used to
extract the solution to measure soluble salts (Jackson,
1958) and was then dried and recovered; thus, subsequent
analyses were performed on samples without soluble
salts. (3) A 0.25-g split was used to analyze total
carbonate (Chittick procedure in Singer and Janitzky,
1986). This split, free of carbonate after the analysis,
was recovered and used to analyze for major oxides and
zirconium. (4) When sufficient sample (0.4 g) existed to
obtain grain size using the Sedigraph rather than by
pipette analysis, the clay and silt fractions were saved
and used to analyze mineralogy by X-ray diffraction.
Most of the laboratory analyses were performed in the
Sedimentation Laboratory of the Institute of Arctic and
Alpine Research in Boulder, Colorado, using standard
laboratory techniques for soil samples (see Black, 1965,
and Singer and Janitzky, 1986) that we adapted for use on
very small samples (the non-organic content of a dust
sample collected from one trap typically weighs less than
1 g/yr). These adaptations generally result in larger
standard errors than normal for the results of different
techniques because the amount of sample used is smaller
than the recommended amount.
Process_Date: 1985
Process_Step:
Process_Description:
Total dust flux is calculated by multiplying the mineral
weight times the fraction less than 2 mm times the pan
area times the fraction of year during which the sample
accumulated (in file labdust.xls, number of days divided
by 365). Other dust-flux values for various components (i.
e. silt flux) are calculated by multiplying the total dust
flux by the percentage of the component.
Preliminary examination of the flux data indicated that
samples from some sites collected in 1985 and 1986, before
the trap design was modified to discourage birds from
roosting, were anomalously large (50-500% greater)
compared to those collected in later years. All of the
anomalous samples had been recorded as having significant
amounts of bird feces at the time of collection.
Consultations with bird biologists confirmed that bird
droppings can contain significant amounts of mineral
matter, mostly derived from cropstones; the amount varies
with the species and with the diet of local populations of
individual species. Moreover, perching birds can
contaminate the sample with material from their feet. In
some cases, we have evidence of near-deliberate
contamination in the form of one or two pebble-sized
clasts of local rocks that were found in samples, possibly
dropped (or swapped for marbles) by large birds such as
ravens. Data from samples with large amounts of bird
droppings were discarded from further analysis and were
excluded from the computations of "selected average" flux
values.
Process_Date: 1987
Process_Step:
Process_Description:
Major elements were measured in U.S. Geological Survey
laboratories on a split of the less-than-2mm fraction
remaining after analysis and removal of carbonate by the
Chittick method. Major elements and zirconium were
analyzed by induction-coupled plasma spectroscopy (Lichte
and others, 1987). In some cases, samples from different
years at the same site or adjacent sites were combined to
obtain enough material for measuring major-oxide
composition.
Process_Date: 1988
Process_Step:
Process_Description:
Major oxides are calculated from elemental compositions
(file dusticp.txt) using the following equations based on
atomic weights:
SiO2 = Si/0.467
Al2O3 = Al/0.529
Fe2O3 = Fe/0.699
MgO = Mg/0.603
CaO = Ca/0.715
Na2O = Na/0.742
K2O = K /0.830
TiO2 = Ti/0.599
MnO = Mn/0.774
ZrO2 = Zr/0.740
The percentages of major oxides and zirconium were then
recalculated to 100%, excluding water, volatiles, and
minor elements, and the ratios of major oxides to ZrO2 are
based on the recalculated values.
Process_Step:
Process_Description:
Mineralogy was measured in U.S. Geological Survey
laboratories on splits of samples that had been previously
analyzed for grain size. Samples of sand, silt, and clay
were slurried in water (sand samples were ground to a fine
powder) and mounted dropwise on glass slides. Minerals
in the sand and silt fractions were identified by
characteristic peaks on X-ray diffractograms and their
relative amounts were estimated by measuring peak
heights. Minerals in the clay samples were identified by
characteristic peaks obtained after the following
treatments: air-dried, glycolated, and heated to 300
degrees C and 550 degrees C. The relative abundances of
clay minerals were estimated by measuring the following
peak heights (in degrees 2 theta) and adjusted for
intensity variations between runs using the peak height of
quartz at 26.65 2 theta: chlorite, 6.3 on the 550 degrees
C trace; kaolinite, 12.6 on the glycolated trace minus the
amount of chlorite; mica, 8.8 on the glycolated trace;
smectite, 5.2 on the glycolated trace; mixed-layer mica-
smectite, 8.85 on the 550 degrees trace minus the amounts
of mica and smectite.
Process_Date: 1988
Process_Step:
Process_Description:
The National Climatic Data Center no longer publishes mean
climatic data for the entire length of record at weather
stations. To obtain mean annual temperature (MAT) and
precipitation (MAP) for the weather stations nearest the
dust traps, averages had to be computed from climatic
summaries of the United States (U.S. Department of
Commerce, 1952, 1965), from station normals for 1961-1990
(National Climatic Data Center, 1992), and from various
climatological data annual summaries. Comparisons could
then be made of the long-term averages with those for the
five years of dust collection (file climate.xls).
Process_Date: 1993
Source_Used_Citation_Abbreviation: DOC 51-60 CA
Source_Used_Citation_Abbreviation: DOC 51-60 NV
Source_Used_Citation_Abbreviation: NCDC 61-90 A CA
Source_Used_Citation_Abbreviation: NCDC 61-90 M CA
Source_Used_Citation_Abbreviation: NCDC 61-90 A NV
Source_Used_Citation_Abbreviation: NCDC 61-90 M NV
Process_Step:
Process_Description:
The dust-trap sites are at different elevations from the
nearest weather stations. To estimate mean annual
temperature (MAT) and precipitation (MAP) at the sampling
sites, annual climate data for the entire period of record
was obtained for every weather station in the region,
including some that are no longer maintained but excluding
those in coastal California. The data in this file was
combined from the data in file aveclim.xls, which included
the weather stations nearest the traps, and from climatic
data for other stations. For many stations with
relatively complete records, this involved computation of
the averages of MAT and MAP (columns under "MAT
calculations" and "MAP calculations") compiled from
records prior to 1961, the last year in which averages for
the entire length of record were published by the U.S.
Department of Commerce (1965), and from station normals
for 1961-1990 (National Climatic Data Center, 1992).
Normals and averages are not published for stations with
missing data or those which were moved at some time; for
these stations, the computation required hand-entering
data for each year of record from the climatological data
annual summaries (columns under "MAT records" and "MAP
records").
Linear regression (bottom left of file) was used to obtain
equations that relate temperature and precipitation to
elevation for these weather stations (columns
"Elevation", "MAT", and "MAP") and to estimate these
parameters at sampling sites with different elevations.
For temperature, only one equation was required; it
provides estimates with a standard error (s.e.) of only
1.3 degrees C. For precipitation, equations were most
useful when the stations were divided into three
geographic regions, including the area of the Mexican
border and the Colorado River-southeast Nevada corridor
(s.e.=2.6 cm), southwestern California east of the
Transverse Ranges (s.e.=8.6 cm), and the interior deserts
(s.e.=2.0 cm).
Process_Date: 1993
Source_Used_Citation_Abbreviation: DOC 51-60 CA
Source_Used_Citation_Abbreviation: DOC 51-60 NV
Source_Used_Citation_Abbreviation: NCDC 61-90 A CA
Source_Used_Citation_Abbreviation: NCDC 61-90 M CA
Source_Used_Citation_Abbreviation: NCDC 61-90 A NV
Source_Used_Citation_Abbreviation: NCDC 61-90 M NV
Process_Step:
Process_Description:
Estimates of MAP and MAT listed under "this study" were
obtained using the linear regression equations calculated
from data in file regclim.xls. These equations are:
MAT = -0.0072E+23.4
MAP (interior deserts) = 0.00555E+7.075
MAP (Colo.R.-Salton Sea) = 0.01013+7.468
MAP (SW Calif.) = 0.05E+5.002
where E is elevation in meters. For comparison, MAP is
also calculated using other published equations. For
stations on the Nevada Test Site (T-1 through T-9) I used
the equation of Quiring (1983), in which y = MAP in inches
and x = elevation in thousands of feet:
y = 1.36x - 0.51
For stations in southern Nevada, including the Nevada Test
Site, I used the equations of French (1983), in which
y = MAP in inches and x = elevation in feet. French (1983)
divided southern Nevada roughly into thirds based on the
paths of moisture-carrying air masses from the west and
south; the eastern third has the most rainfall, the
western third has the least, and the central third is
intermediate:
Eastern: log y = 0.0000933x + 0.486
Central: log y = 0.0000786x + 0.446
Western: log y = 0.0000365x + 0.505
MAP at the closest weather station to the dust-trap site
is also given. Estimates of MAP for sites near Los
Angeles, including T-51 through T-54, using the equations
from this study gave unrealistically low values (see file
trapclim.xls) because this area is under a coastal rather
than an interior climate. Thus, in the papers written
using these data, MAP for these sites is assumed to be
about the same as that at the nearest weather station.
Process_Step:
Process_Description:
Mean monthly precipitation and temperature from 1984 to
1989 were acquired from the National Climatic Data Center
(1984-1989) for weather stations in southern Nevada and
California that were closest to dust-trap sites and
entered into a spreadsheet in order to calculate mean
annual values for climatic variables and compare them to
long-term means (calculated in file aveclim.xls).
Seasonal precipitation (May-October and November-April)
was calculated from monthly values.
Process_Date: 1993
Process_Step:
Process_Description:
Secondary climatic variables were calculated from the data
in file climate.txt These secondary variables include
monthly and annual potential evapotranspiration (PET) and
the leaching index (LI) of Arkley (1963). The leaching
index is a measure of available moisture obtained by
subtracting monthly evapotranspiration from monthly
precipitation. PET was calculated for all stations with
both temperature and precipitation data using the method
of Thornthwaite (1948), and for stations with mean
minimum and maximum temperatures using the method of
Papadakis (1965). The leaching index is calculated for
both methods of PET. Pan evaporation measurements are
also given where available (National Climatic Data Center
and Farnsworth and others, 1982) for comparison.
PET is more readily calculated by the Thornthwaite method
than by the Papadakis method, because the latter requires
mean minimum and maximum temperatures that are commonly
not recorded at some weather stations. However, according
to Taylor (1986), the Thornthwaite method applied to
climatic data for arid regions yields PET values that are
much too low (as much as 150% compared to evaporation-pan
data for the growing season). The Papadakis method
provides estimates of PET that are closest to pan data in
arid climates. Many thanks to Emily Taylor (U.S.
Geological Survey) for guiding me through the complex
calculations of PET and providing me with the appropriate
references.
[Editor's note: These equations contain expressions that
cannot be conveniently represented in plain ASCII text.
Accordingly, I have coded the expressions using the
notation of the programming language BASIC, hoping that
most people will understand that. BASIC has no
subscripting, however, so I used the underscore to
indicate that the next character or two is subscripted.
The correct notations can be obtained by examining the
original document, in Microsoft Word for DOS format.]
LI = (P - PET) summed for each month in which P > PET.
PET (Thornthwaite) = F(1.6(10t/I)^a)
where
t temperature (degrees C) for the month
I sum for 12 months of (t/5)^1.514 (given in column "heat factor I")
a (6.75*10^(-7) * I^3) - (7.71* 10^(-5) * I^2) + (0.1792 * I) + 0.49239 (given in column "exponent a")
F day length factor (from table V in Thornthwaite, 1948)
PET (Papadakis) = 5.625 (e_ma - e_d)
where
e_ma is saturation vapor pressure of monthly average daily maximum temperature (mbars)
e_d is monthly average vapor pressure (dew point) (mbars)
According to Lindsley and others (1975, p. 35), vapour
pressures are calculated by:
e_ma = (33.869(0.00738 (max.T) + 0.8072)^8 - 0.00019 |1.8 (max.T) | + 0.001316)
e_d = (33.869(0.00738 (min.T) + 0.8072)^8 - 0.00019 |1.8 (min.T) | + 0.001316)
where max.T is the monthly average maximum temperature and
min.T is the monthly average minimum temperature.
[Editor's note: Here are the preceding equations rendered
in TeX:
{\parskip=\medskipamount
$LI = (P - PET)$ summed for each month in which $P > PET$.
$$PET (\hbox{Thornthwaite}) = F(1.6(10t/I)^a)$$
where
$$\halign{\quad # \hfil & \quad # \hfil\cr
$t$ & temperature (degrees C) for the month\cr
$I$ & sum for twelve months of $(t/5)^{1.514}$ (given in column ``heat factor I'')\cr
$a$ & $(6.75 \times 10^{-7}I^3) - (7.71 \times 10^{-5}I^2) + (0.1792I) + 0.49239$ (given in column ``exponent a'')\cr
$F$ & day length factor (from table V in Thornthwaite, 1948)\cr
}$$
$$PET (\hbox{Papadakis}) = 5.625 (e_{ma} - e_d)$$
where
$$\halign{\quad # \hfil & \quad # \hfil\cr
$e_{ma}$ & is the saturation vapor pressure of monthly average daily maximum temperature (in mbar), and \cr
$e_d$ & is the monthly average vapor pressure (dew point) in mbars\cr
}$$
According to Lindsley and others (1975, p. 35), vapor pressures are calculated by:
$$e_{ma} = (33.869(0.00738 (\hbox{max.T}) + 0.8072)^8 - 0.00019 \vert 1.8 (\hbox{max.T}) \vert + 0.001316)$$
$$e_d = (33.869(0.00738 (\hbox{min.T}) + 0.8072)^8 - 0.00019 \vert 1.8 (\hbox{min.T}) \vert + 0.001316)$$
where max.T is the monthly average maximum temperature and min.T is the monthly average minimum temperature.
}
[Editor's note: end of TeX rendition of the equations.]
