3. Climate of Arizona
Ranges of temperatures and precipitation in Arizona are extreme. Average annual temperatures vary from the middle 20s C (70s F) in the low desert areas along the Gila and Colorado rivers to less than 5 C (middle 40s F) in the pine country of central and east-central Arizona. And average annual precipitation ranges from about 76 mm (3 in) in the southwest to more than 760 mm (30 in) in the central and eastern mountains. A general schematic of atmosphere-to-soil precipitation distribution is shown in Figure 27.
Record high and low temperatures span nearly 71 C (167 F). An all-time low of -40 C (-40 F) was recorded January 7, 1971, at Hawley Lake in the White Mountains near McNary. The record high of 52.8 C (127 F) was set twice, once at Fort Mohave on June 15, 1896, and again at Parker on July 7, 1905. Both towns are along the lower Colorado River. In fact, summer temperatures above 49 C (120 F) have been reported at all towns along the Colorado River south of Hoover Dam and along the Gila River west of its confluence with the Salt River.
The state's warmest and coldest towns are, respectively, Mohawk, about 80 km (50 mi) east of Yuma, and Maverick in the southwest corner of Apache County on the Fort Apache Indian Reservation. Average July temperatures at these towns differ by almost 20 C (35 F), the average at Mohawk being 34.6 C (94.3 F) and at Maverick 15.5 C (59.9 F). Arizona's range of thermal climates, then, varies from the long, hot summers and short, mild winters of the warm deserts, to the short, cool summers and long, icy winters of the cold highlands.
This wide range of climatic conditions in Arizona is due to four important factors:
FIGURE 27. Schematic of Precipitation Distribution from Atmosphere to Soil (after E. M. Bridges, 1978)
The great range of mean annual air temperatures in Arizona is mostly due to elevation. This was illustrated by Smith ( 1956) who plotted the mean annual air temperatures of selected Arizona weather stations as a function of elevation. The mean January and mean July air temperatures are strongly influenced by elevation also (Plates 4, 7 and 8). The coldest mean January air temperature of -7 to -4 C (20 to 25 F) are on the Kaibab Plateau, in the White Mountains and in the San Francisco Peaks. The warmest mean January air temperatures are greater than 13 C (55 F) along the lower Gila River.
Latitude also appears to affect temperatures since the weather stations in northeastern Arizona have mean January air temperatures about 5.5 C (10 F) colder than those at comparable altitudes in the southeast. Mean July air temperatures, on the other hand, drop rather uniformly with increased elevation. Temperatures decrease about .55 C (1 F) per 71 m (235 ft) increase in elevation (Sellers and Hill, 1974).
Roughly half of Arizona receives less than 250 mm (10 in) average annual precipitation. Areas in this regime are the southwestern, extreme western and a large part of the northeastern portions of Arizona as well as most of the San Simon Valley in the southeast. The wetter parts of Arizona receive up to 635 mm (25 in) or more average annual precipitation and are at the higher elevations along the Mogollon Rim, the southern part of the Kaibab Plateau, the White Mountains, the San Francisco Peaks and several mountains in the Basin and Range Province (Plate 5).
A unique feature of Arizona climate is the two periods of precipitation; one season is from December through March and the other during July, August and September. The proportion of winter precipitation decreases from west to east and the summer precipitation decreases from south to north (Plate 6). The biseasonal precipitation pattern appears to reach equality in the Flagstaff-Prescott area with half the average annual moisture falling during each period (Jameson, 1969).
Winter precipitation is the ‘‘Mediterranean’’ type, the same as that which provides the strong winter maximum moisture in California. The storms are associated with large-scale, mid-latitude cyclonic disturbances. Arizona generally receives only the fringes of these storms. Occasionally the semipermanent ridge of high pressure in the Pacific Ocean that is normally off the West Coast is displaced westward and a low pressure trough from the east is drawn over the western United States. When this happens, storms, rather than moving inland farther north, enter central and southern California and move inland across Arizona (Jurwitz, 1953). This weather pattern produced the wettest winters recorded in Arizona (Sellers and Hill, 1974). The maximum moisture from these storms falls on the central portion of the Mogollon Rim. Downwind from the Rim is a distinct rainshadow that is especially pronounced in the Little Colorado River Basin. The result is that the Mogollon Rim has a winter wet climate on the windward, or west, side and a winter dry climate on the lee, or east, side.
Summer rainfall has a tropical origin. The influx of warm, moist air masses from the Gulf of Mexico begin about July 1 and last until the end of August. When the Pacific high pressure cell off the West Coast moves rapidly northeast in late June, the Southwest receives a deep, gentle air flow from the Gulf of Mexico on the southwest side of a high pressure cell that protrudes from the Atlantic Ocean into the central part of the United States (Bryson and Lowry, 1955; Bryson, 1957). The unstable air advances into Arizona from the southeast over strongly heated land surfaces and yields moderate to heavy afternoon or evening thundershowers. This convective activity is most pronounced near mountains (Sellers and Hill, 1974). This phenomenon has been referred to as a monsoon (Huntington, 1914).
Summer rains are brief, sometimes intense, particularly when the storm begins, and scattered, seldom affecting more than several square kilometers (a few square miles) (Humphrey, 1933; Osborn, 1983). These storms tend to occur in a region for several days, then several days of dry weather follow. Monsoon air masses do not show distinct frontal characteristics and no rainshadow effect develops, as with winter storms. Areas where topography changes abruptly receive more rain. As it is in winter, summer precipitation is heaviest on the central Mogollon Rim (Jameson, 1969).
Hurricane influences also appear in Arizona and may contribute to the late summer rainfall. These hurricanes are not from the Caribbean, but from off the west coast of Mexico and occasionally drift up into northern Mexico and southern Arizona. Although they lose surface circulations characteristic of true hurricanes as they move inland, they do bring deep, moist air masses that have given the Southwest and northern Mexico some of the heaviest rainfalls on record (Thorud and Ffolliot, 1973; Sellers and Hill, 1974). These storms, when they occur, are in late August and September.
Records at Arizona climate stations show that precipitation is quite variable from year to year, but arrival time and expected amount of summer precipitation are less variable than those of winter storms at most stations. McDonald ( 1956) computed coefficients of variation of winter, summer and yearly amounts of precipitation for Yuma, Phoenix, Tucson and Flagstaff (Table 3). The coefficients of variation
|Yuma||0.75 ± .06||0.94 ± .07||0.62 ± .05|
|Phoenix||0.57 ± .05||0.56 ± .05||0.40 ± .03|
|Tucson||0.54 ± .04||0.40 ± .03||0.30 ± .02|
|Flagstaff||0.37 ± .04||0.28 ± .04||0.23 ± .02|
|Cw--Coefficient of variability of winter precipitation|
|Cs--Coefficient of variability of summer precipitation|
|Cy--Coefficient of variability of yearly precipitation|
represent the ratios of the standard deviation amounts to the mean amounts. The unusually high coefficient of variation of summer precipitation for Yuma is due to the extremely low amounts of summer precipitation. At Tucson and Flagstaff the coefficients of variation of summer precipitation clearly are lower than the variations of winter precipitation.
Arizona has three basic climatic regimes based on average annual precipitation and mean temperature. Desert regimes cover about 30 percent of Arizona, the steppes about 53 percent and the highlands about 17 percent. These three regimes are each subdivided further into warm and cold, depending on whether the average temperature of the coldest month is above or below 0 C (32 F) and are illustrated in Figure 28. (The reader is invited to compare these regime boundaries with those of other plates and figures in this book.)
Deserts and steppes are characterized by limited amounts of precipitation; the former more so than the latter. Practically all of the moisture that falls in these regions evaporates; appreciable runoff and subsurface storage seldom occur (See Figure 29 and Plates 5 and 13). As result, the vegetation cover consists mostly of creosotebush, cacti and sagebrush on the deserts and mesquite, pinyon-juniper and various grasses on the steppes (see Plates 11 and 12). Irrigation is a must for successful farming in these dry regions. On the other hand, the highlands, particularly the cold highlands, normally receive sufficient precipitation during the year to support moderately dense vegetation and to provide substantial runoff to the surrounding drier areas (See Figure 30 and Plates 5 and 13). Some of the finest and most extensive pine forests in the world are in Arizona's cold highlands where precipitation is reasonably dependable from year to year. Precipitation is less dependable in the warm highlands, varying greatly in amount and intensity from year to year.
Summer rains are heaviest and most dependable in the highlands. A dry afternoon between the second week of July and the first week of September is rare. Occasional cloud-bursts send water raging down into the surrounding valleys, filling washes and arroyos to overflowing and doing considerable damage to roads and poorly sited buildings. These storms also occur over desert towns, where they can be especially destructive. Sewer systems in most of these communities were not built to handle more than moderate amounts of runoff and are inadequate to channel the large volumes of runoff produced by the intense summer storms.
The driest regions in winter generally are the cold steppes, particularly those north of the Little Colorado River. These areas are ‘‘protected’’ from major sources of winter moisture by high mountain ranges and plateaus including the White Mountains, San Francisco Peaks and the Mogollon Plateau to the south and southwest, the Wasatch Mountains of Utah and the Rocky Mountains to the northeast. These barriers, however, do not block unusually strong winds, particularly during winter and spring. When a storm system passes over the area, winds exceeding 50 kmh (30 mph) may last for several days, first warm from the south, then bitterly cold from the north.
At least a trace of snow has been recorded in all parts of Arizona, but usually only in the mountain sections are snow depths measurable. Seasonal totals vary from 300 to 1,500 mm (1 to 5 ft). Slopes of the higher mountains provide excellent opportunities for winter sports, even those rising from the desert regions. Early season snowstorms are very unpredictable; therefore, it is a good policy for hunters, hikers and campers to be prepared for snow when they go into the mountains in the late fall months. Motorists also should be prepared to encounter unexpected snowfall while driving in the mountains during this season.
The southern Arizona desert has mild, dry winters that are very important in its rapid population growth. Temperatures in the coldest month usually range in the early morning from 0 to 2 C (middle 30s F) to 18 to 20 C (high 60s F) in the midafternoon. Below-freezing temperatures are rare. When they do occur they are limited for the most part to the time just before and during sunrise, and to low areas that receive drainage of cold, dense air from higher surrounding terrain during the night.
Many persons find Arizona desert summers less uncomfortable than those of the eastern United States despite the high temperatures. The extreme dryness of the air, with relative humidities often falling below 10 percent in May and June, relieves intensity of the heat by increasing evaporation, a mechanism that helps cool the body. And virtually all buildings are cooled either by refrigeration or by evaporative coolers. When humidity is low even the least expensive evaporative coolers are quite effective. But when humidities are high, particularly during the July- August rainy season, evaporative coolers are rather ineffective.
Arizona's warm steppes are slightly more cool, wet and humid throughout the year than the warm deserts but they still generally are characterized as being hot and arid. These regions are close to the highlands that provide refreshingly mild days and cool nights during the hot summer months. The warm steppes, particularly those in southern Arizona, are popular centers of winter recreational activities. The temperature variation between day and night in the cooler months often exceeds 22 C (40 F). The early morning minimum is between -2.8 and -1.7 C (upper 20s F) and the afternoon maximum between 21.5 and 23 C (low 70s F).
Perhaps the most rugged climate is in the Arizona cold steppe regions that are almost entirely in the northeast. Winters are cold, dry and windy; the temperature normally falls below -17.8 C (0 F) on four to six days between November and March. Summers are quite warm, especially in the northern regions, where summer afternoon temperatures above 32 C (90 F) may be expected 50 to 65 days. But summer nights are cool and temperatures usually fall to the 10s C (50s F) by sunrise.
FIGURE 28. Climates of Arizona (after UA Press, 1972)
FIGURE 29. Generalized Water Cycle in Hyperthermic, Thermic and Some Mesic Regions
FIGURE 30. Generalized Water Cycle in a Frigid Subhumid Region
The Arizona General Soil Map mapping units are grouped into seven climatic zones. These zones are a combination of the four soil temperature regimes and three precipitation zones, arid, semiarid and subhumid. The seven climatic zones are defined in Table 4.
It has been shown that mean annual soil temperature is rather closely related to mean annual air temperature (Smith et al, 1964). Factors that may affect this relationship include amount and distribution of rain and snow, shade and litter layers in forests, percent and direction of slope and irrigation. Methods for calculating soil temperature with multiple regression equations by using site factors such as mean annual air temperature, elevation, latitude, slope, aspect and certain soil characteristics have been proposed (Arkley, 1971; Ouellet, 1973; Konstantinov and Popovich, 1980; Meikle and Treadway, 1981).
Soil temperature is an important property that controls or has a strong influence on plant growth and soil formation. Every pedon has a mean annual temperature that is essentially the same at all depths, but at any given moment the temperature of any two horizons is rarely the same because of daily, short-term and seasonal fluctuations (Smith et al, 1964). Soil temperature data for Arizona are limited. U.S. Forest Service soil scientists, however, recently have started taking soil temperature measurements under a variety of environmental conditions. In some cases several years of data are available (Robbie, Brewer and Jorgensen, 1982).
Each pedon has a characteristic temperature regime that can be measured and described. For most practical purposes, the regime can be described by the mean annual soil temperature, the average seasonal fluctuations from that mean, and the mean warm or cold seasonal soil temperature gradient at depths from 5 to 100 cm (2 to 40 in), the main root zone (Soil Survey Staff, 1975).
|Symbol||Unit Name||Soil Temperature *||Precipitation|
|HA||Hyperthermic Arid||22 C (72 F) or greater||250mm (10 in) or less|
|TA||Thermic Arid||15 to 22 C (59 to 72 F)||130 to 250 mm (5 to 10 in)|
|TS||Thermic Semiarid||15 to 22 C (59 to 72 F)||250 to 460 mm (10 to 18 in)|
|MA||Mesic Arid||8 to 15 C (47 to 59 F)||150 to 250 mm 6 to 10 in)|
|MS||Mesic Semiarid||8 to 15 C (47 to 59 F)||250 to 410 mm (10 to 16 in)|
|MH||Mesic Subhumid||8 to 15 C (47 to 59 F)||410 mm (16 in) or greater|
|FH||Frigid Subhumid||8 C (47 F) or less||410 mm (16 in) or greater|
Hyperthermic soils have mean annual soil temperatures of 22 C (72 F) or more and a greater than 5 C (9 F) difference between mean summer (mean temperature of June, July and August) and mean winter (mean temperature of December, January and February) temperatures at a depth of 50 cm (20 in) or at the soil-rock interface in shallow soils, soils less than 50 cm (20 in) deep.
Thermic soils have mean annual soil temperatures of 15 C (59 F) or more, but less than 22 C (72 F), and the difference between mean summer and mean winter temperatures is greater than 5 C (9 F) at a depth of 50 cm (20 in) or at the soil-rock interface in shallower soils.
Mesic soils have mean annual soil temperatures of 8 C (47 F) or more, but less than 15 C (59 F), and a difference between mean summer and mean winter temperatures greater than 5 C (9 F) at a depth of 50 cm (20 in) or at the soil-rock interface in shallower soils.
Frigid soils have mean annual soil temperatures of more than 0 C (32F), but less than 8 C (47 F) and a difference between mean summer and mean winter temperatures greater than 5 C (9 F) at a depth of 50 cm (20 in) or at the soil-rock interface in shallower soils.
Arizona also has cryic soils. These soils, like the frigid soils, have mean annual temperatures between 0 C (32 F) and 8 C (47 F), but they are considered to be colder than the frigid soils since they are colder in summer. Pergelic soils have mean annual soil temperatures less than 0 C (32 F) and may be in Arizona at the highest elevations of the San Francisco Peaks.
Another aspect of the influence of temperature on soils is the number of diurnal freeze-thaw cycles per year. This aspect may be significant in soil formation because of frost weathering (Potts, 1970). The White Mountains have the most days in which the temperature fluctuates across the 0 C (32 F) mark (Figure 31).
FIGURE 31. Frost-Alternation Days in Arizona (after R. K. Merrill and T. L. Péwé, 1977)
Applying the soil moisture regime to classify and map Arizona soils is difficult because of the lack of data. And obtaining data about how moisture in the moisture control section changes with the seasons is difficult also. Consequently less direct methods have to be used to approximate soil moisture regimes. Soil moisture balance calculations based on climatic data as described below commonly are used (Newhall, 1980). Soil moisture and soil temperature regimes also are sometimes inferred from vegetation. Plant communities, or alternatively a few index species, are associated with particular soil moisture and temperature regimes. Since the moisture balance methods have shortcomings and since few studies of natural vegetation and its soil moisture and temperature requirements have been attempted, there is a general lack of agreement on identifying soil moisture regimes, especially in the western United States (Daugherty, 1982). As Daugherty ( 1982) also pointed out, the descriptons of soil moisture regimes in Soil Taxonomy (Soil Survey Staff, 1975) were based mostly on cultivated areas outside the western states and do not seem to apply as well in the West. Finally, the authors of Soil Taxonomy (Soil Survey Staff, 1975) recognized that the definitions of the soil moisture regimes were far from perfect and that revisions would be expected.
Nonetheless, soil moisture regimes can be determined in part by precipitation and evapotranspiration and may be influenced by topography and its effect on runoff, consequently on runon (Plate 13). Soil properties such as texture, organic matter content, type of clay, structure and soil depth as they affect infiltration and retention of moisture also influence soil moisture regimes. Three categories, dry, moist and saturated, are used to define soil moisture regimes (Soil Survey Staff, 1975). Soil is considered dry when moisture tension is greater than 15 bars. Available water in soils that are moist is held between 0 and 15 bars. Soils are saturated when there is no tension on the water and essentially all the void space is filled with water. The depths in the soil at which soil moisture criteria are applied are established by the concept of the moisture control section. The upper boundary of the moisture control section is the depth to which a dry soil will become moist 24 hours after a 25 mm (1 in) rain. The lower boundary is either the depth to which 75 mm (3 in) of water, introduced through the soil surface, will moisten a dry soil in 48 hours, or the depth to a pan, bedrock or other root-restricting layer, whichever is shallower.
Soil moisture regimes described in Soil Taxonomy (Soil Survey Staff, 1975) are based on the time and season that the soil moisture control section is saturated, moist or dry. The definitions of these soil moisture regimes were summarized by Buol, Hole and McCracken ( 1980) and are repeated below.
Aridic or Torric Moisture Regime. These soils are both dry more than half the time when not frozen and never moist more than 90 consecutive days when soil temperatures are above 8 C (47 F) at 50 cm (20 in) depth.
Ustic Moisture Regime. In most years, these soils are dry for more than 90 cumulative days but less than 180 days. In temperate climates, they are usually moist at least 45 consecutive days in the four months after the winter solstice and not dry 45 consecutive days in the four months after the summer solstice.
Xeric Moisture Regime. These soils are only in the temperate areas where summers are dry and winters moist. These soils usually are dry more than 45 consecutive days in the summer and moist more than 45 consecutive days in the winter.
Soils with aridic (torric) and ustic moisture regimes are the most widespread in Arizona. Aridic soils constitute essentially all Hyperthermic Arid (HA), Thermic Arid (TA) and Mesic Arid (MA) soils, and most of Thermic Semiarid (TS) and Mesic Semiarid Soils (MS) shown on the Arizona General Soil Map (Plate 1). Some ustic soils are included in Thermic Semiarid (TS) and Mesic Semiarid (MS) soils. Ustic and Udic soils constitute the Mesic Subhumid (MH) and the Frigid Subhumid (FH) soils with udic soils in the higher rainfall areas. Small areas of aquic soils are in topographically low areas that have poor drainage and limited runoff. Soils with xeric regimes are not recognized in Arizona. Areas in northern Arizona near the Utah border contain soils that might approach having a xeric moisture regime but are considered ustic because they usually receive enough summer rain to prevent the soil moisture control section from becoming dry for a sufficient length of time. Perudic moisture regimes are not present in Arizona soils.
Temperature and precipitation alone are poor descriptors of climate. The amount of precipitation does not indicate whether a climate is moist or dry unless the water need of the site can be compared with it. And temperature does not really reveal the energy that is available for plant growth and development unless the moisture condition of the soil is known. Thus, one of the major objectives of the water balance approach in characterizing climate is to arrive at a better way of determining whether a climate is moist or arid by comparing the climatic moisture supply with the moisture needs. Water balance estimations also have been quite useful for correlating some climatic characteristics of a region that are important in soil formation (Hurst, 1951; Arkley, 1963, 1967), in soil classification (Cox, 1968), in plant growth (Arkley and Ulrich, 1962), in distribution of vegetation (Mather and Yoshioka, 1968) and in determination of soil moisture regimes (Newhall, 1980).
Thornthwaite's ( 1948) procedure makes it possible to estimate soil moisture conditions of a site from the gains and losses of soil moisture over a certain interval of time, day, week or month. Gains are from precipitation, the data of which are obtained from weather stations. Losses are evapotranspiration, the combined loss from the ground and plant surfaces (evaporation) and the loss of water from living plants (transpiration). Potential evapotranspiration is the amount of moisture lost by soil when it is covered by vegetation and amply supplied with water. Thornthwaite ( 1948) studied water-use rates in irrigation projects and catchment runoff records. He discovered a close relationship between mean monthly temperature and potential evapotranspiration if adjustments are made for variations in day length.
From the findings of these studies Thornthwaite ( 1948) developed a rather complicated formula for computing potential evapotranspiration of a site if the latitude is known and temperature records are available. The computations involved are simplified by the use of tables (Thornthwaite and Mather, 1957) and/or graphs (Palmer and Havens, 1958). Potential evapotranspiration differs from the actual evapotranspiration (see Plates 9 and 10). The actual water loss to the atmosphere equals potential evapotranspiration only during those periods when precipitation is greater than the potential evapotranspiration. When precipitation is less than potential evapotranspiration, actual evapotranspiration is equal to the sum of precipitation and water lost by soil to the atmosphere.
When precipitation is equal to water need or potential evaporation, the soil is at field capacity and no leaching takes place. When the soil water supply is greater than the need, there is a surplus of moisture for leaching the soil and subsequent lowering of the soil pH over time. When it is less, the soil moisture in the root zone is drawn upon by vegetation until the wilting point is reached and the moisture budget becomes deficient. After the dry season, the soil-moisture reservoir must be replenished to field capacity before there can be a surplus again.
Thornthwaite ( 1948) originally assumed that except in shallow soils, soils at field capacity have 10 cm (4 in) of water available for the atmosphere and vegetation to draw on if precipitation does not fall. Arkley ( 1967) and Steila ( 1972) suggested 15 cm (6 in) as a better value of soil-moisture storage capacity. Thornthwaite and Hare ( 1955), however, suggested that at least 30 cm (12 in) of water are available to deep-rooted, mature plants in most soils. Thornthwaite ( 1948) also originally assumed that all stored moisture, 10 cm (4 in), was equally available. However, it is well known that as soil dries, it becomes increasingly difficult for water to be lost to evaporation and transpiration. Thus, as the soil-moisture content decreases, so too does the rate of evapotranspiration. Thornwaite and Mather ( 1955), Baier and Robertson ( 1966) and Baier ( 1969) suggested different relationships between actual evapotranspiration and soil-moisture content. Thornthwaite and Mather ( 1955) assumed a linear relationship between water loss and soil-moisture content, meaning that when the soil moisture is reduced to 50 percent of capacity, the actual evapotranspiration rate is only 50 percent of the potential rate.
Baier ( 1969) measured moisture of soils used for dryland wheat cultivation in Canada for 10 years. He found essentially no reduction in soil moisture depletion rates from 100 percent to 70 percent available water. With further soil drying, an approximate linear relationship was observed between moisture loss and moisture content from 70 percent to 0 percent available water.
It is now recognized that the Thornthwaite calculation method underestimates potential evapotranspiration in arid regions. This was illustrated by Gay ( 1981) who compared potential evapotranspiration estimates for Tucson by several methods. The Thornthwaite method produced the lowest values when compared with other calculation methods and with pan and adjusted pan evaporation rates.
Average annual water budgets of 14 weather stations representative of the seven temperature-precipitation zones outlined in Table 4 are illustrated in Figure 32. The potential evapotranspiration values for each month were calculated by the Thornthwaite ( 1948) method using Palmer-Havens (Palmer and Havens, 1958) graphics techniques. A soil-moisture storage capacity of 10 cm (4 in) and a linear relationship between water loss and soil moisture content were assumed (Thornthwaite and Mather, 1955). Climatic data used in the calculations are from Sellers and Hill ( 1974). The annual water budgets depicted in Figure 32 should be considered to be qualitative representations of changes in soil-moisture conditions since the estimates of potential evapotranspiration may be too low, especially for the more arid stations. Potential evapotranspiration values and water budget tabulations for other Arizona stations were compiled by Buol ( 1964).
Yuma, representative of the Hyperthermic Arid region, is one of the hottest and driest places in Arizona as illustrated in Figure 32. The potential evapotranspiration is greater than precipitation every month and the soil-moisture budget is deficient throughout the year. Phoenix, also in the Hyperthermic Arid region, receives slightly more moisture than Yuma that provides some soil-moisture recharge during the winter and some soil-moisture use in the early spring. All other stations, except Fort Valley and Maverick in the Frigid Subhumid region, have soil-moisture recharge during winter and soil-moisture use and deficit during spring and summer. The amount of moisture available during spring is higher at stations that have more winter precipitation and lower temperatures. Thus, the soil-moisture recharge during winter is greater at Kingman than at Lees Ferry, at Bagdad than at Douglas, at Tuweep than at Springerville and at Whiteriver than at Canelo. The differences between soil-moisture budgets at stations in similar climatic regions are due to
FIGURE 32. Generalized Annual Moisture Budgets for Soils near 14 Climate Stations in Arizona (Format after M. E. Hecht and R. W. Reeves, 1981, based on C. W. Thornthwaite and J. R. Mather, 1957)
Many stations, particularly those in southeastern Arizona, receive more than 50 percent of the average annual precipitation during July, August and early September when potential evapotranspiration is high. In spite of moderate to high amounts of precipitation, therefore, the water budget method may forecast fairly large soil-moisture deficits, indicating that precipitation and water stored in the soil are inadequate to meet water needs.
Actual evapotranspiration may be fairly high during the summer rainy season, but lower than potential evapotranspiration because of vegetative growth during July, August and September in the subhumid, semiarid and to a lesser extent in the arid regions of Arizona. This observation supports the contention that the soil-moisture deficit during the summer rainy season in Arizona is misleading as calculated by Thornthwaite's ( 1948) water budget technique.
It is emphasized that water budgets shown in Figure 32 are based on averages. Since precipitation at the climatic stations varies considerably from year to year, water budgets also vary from year to year. Extremely wet seasons, such as the winters of 1973 and 1980 in Arizona, produce greater excess soil moisture for leaching and/or recharge of soil moisture for use in the spring. Dry years, on the other hand, have less intense leaching and less moisture for recharge.
Climatic conditions in the past, however, may have been more influential than those of more recent times in determining many properties of Arizona soils. Knowledge of past climatic conditions, therefore, is highly desirable in interpreting the influence of climate on soil properties and gaining a greater understanding of the relationships between soils.
A wealth of evidence is accumulating from which generally consistent inferences about the climate of Arizona during the past 50,000 years or so can be made. This evidence comes chiefly from paleobotanical, palynological, dendrochronological, paleozoological, glacial and pluvial studies in the Southwest.
Pleistocene alpine glaciation in Arizona occurred in the San Francisco Peaks near Flagstaff (Sharp, 1942; Péwé and Updike, 1970; Updike and Péwé, 1974; Duncklee, 1978) and in the White Mountains near Mount Baldy (Melton, 1961; Merrill and Péwé, 1972, 1977). The glaciations are evidence of colder climatic conditions, but only very small areas in mapping unit FH2, Sponseller-Ess-Gordo Association, were directly affected.
Periglacial phenomena occurred in nonglaciated high elevations during glaciation on Kendrick Peak near Flagstaff (Barsch and Updike, 1971) and in the Chuska Mountains (Blagbrough, 1971), and were suggested by Melton ( 1965) in the Pinaleno and other high mountains in southern Arizona. Galloway ( 1970) and Melton ( 1965) indicated that physical weathering associated with periglacial environments in the mountains of the Southwest was important in providing rock and other coarse fragments common in many colluvial and alluvial fan deposits and soils.
During periods of the Pleistocene many extensive and relatively deep lakes filled basins in the Southwest that now are dry or contain shallow, saline lakes (Feth, 1961). Pluvial has been used to describe the times when the lakes existed and implies that precipitation was somewhat greater than now. Most authorities believe that the more moist climatic conditions during glacial-pluvial times were caused by the combined effects of increased precipitation and decreased potential evapotranspiration due to lower temperatures.
Biotic communities in Arizona were displaced about 300 m (1,000 ft) or more to lower elevations and hundreds of kilometers (miles) south during glacial-pluvial times. Evidence of this displacement consists mostly of pollen stratigraphy and plant materials preserved in fossil packrat middens.
A number of pollen-stratigraphic studies made in the Southwest were reviewed by Hevly and Karlstrom ( 1974). The results of these studies showed that the Southwest paleoclimate was in phase with those of the Pacific Coast and mid-continental North America. When continental glaciers were expanding, Southwest biotic communities were displaced to lower elevations and more southerly latitudes. Pluvial lakes expanded and alluvial deposition was augmented by the actions of more moist and cooler climates. As continental glaciers waned, the reverse biotic, geologic and climatic phenomena occurred.
Widespread occurrences and discoveries in the Southwest of extremely old packrat (genus Neotoma) middens containing abundant, well-preserved plant macrofossils recently have provided good information about the nature of paleoplant communities and inferred climatic conditions (Lanner and Van Devender, 1974; Wells, 1976). Analyses of fossil plant remains yield more reliable information than pollen analyses. Packrats thoroughly sampled vegetation within 100 m (330 ft) of the midden sites. But with pollen stratigraphy there is always the possibility that some pollen was introduced to a site from some distance away. Analyses of plant remains also often allow for more specific identification in certain genera. Radiocarbon ages of middens from a number of sites in the Mohave, Sonoran and Chihuahuan deserts have been reported to range in age from less than 4,000 to 40,000 years. The results of all of these studies show vegetation changes indicative of cooler and/or more moist climates in the three deserts during the late Pleistocene.
Based on interpretations of packrat midden data and to a lesser extent on pollen stratigraphy and geologic evidence of snowline lowering, three principal models were developed of climatic conditions during the glacial maximum of the late Wisconsin in the Southwest.
In the second climatic model, Wells ( 1979) disagreed with the Brakenridge ( 1978) and Galloway ( 1970, 1983) interpretations. He believed that the glaciopluvial climates of the Southwest desert lowlands had more equable temperatures. Wells ( 1979) also suggested considerably more summer rainfall than now, but with a strong southeast to northwest gradient of decreasing rain with increasing distance from the Gulf of Mexico, similar to the present gradient. Thus, the Mohave Desert to the northwest was relatively cool and dry and the Chihuahuan Desert to the southeast was relatively warm and moist. Wells ( 1979) also did not believe that winter precipitation increased although its effectiveness would have been enhanced by cooler temperatures.
The third climatic model proposed by Van Devender ( 1977) and Van Devender and Spaulding ( 1979), disagreed with Wells ( 1979). They believed that the late Wisconsin climate of the Southwest was characterized by more winter precipitation than now, probably because of more numerous frontal storms south of the crest of the Sierra Nevada (36°N) that extended as far east as the Trans-Pecos in Texas. These frontal storms also may have begun earlier in the fall than now and lasted later in the spring. Van Devender and Spaulding ( 1979), also suggested that the late summer-early fall hurricanes were moved farther south by colder sea surface temperatures so that that source of precipitation was eliminated from the Southwest during the late Wisconsin. Moreover, the winter temperatures, according to Van Devender and Spaulding ( 1979), were rather mild, perhaps comparable with the present, while the summers were cool. The mild winters could have been due to the thick continental ice sheets that altered atmospheric air currents, drawing warmer currents from the Pacific and preventing cold Arctic air masses from entering the mid-continent and the Great Basin areas. Cole's ( 1982) packrat midden data also indicated that precipitation in the eastern Grand Canyon fell predominantly in the winter but that the Wisconsin glacial climate was colder in all seasons than it is today. Cole's ( 1982) data suggested a wider range between summer and winter mean temperatures.
The elevational lowering of vegetation zones implied by Hevly and Karlstrom ( 1974) is not as straightforward as it might seem. Brakenridge ( 1978) suggested the telescoping effect. The upper elevational limit of a given vegetation zone was lowered more than the lower boundary. He indicated that this effect was most marked on ponderosa pine and severely restricted its vertical range. Cooler summers depressed the upper boundary, while expansion to lower elevations was limited by soil moisture deficits. Cole ( 1982) found evidence that glacial vegetation of the Grand Canyon was similar to modern vegetation in northern Utah and that the composition and elevational zones of the plant associations changed. Thus, Cole ( 1982) concluded that Pleistocene vegetational zones were not simply depressed versions of modern zones but rather reflected a latitudinal shift of climate in the Grand Canyon region.
One difficulty in evaluating climatic conditions during glaciation is that lesser glaciations with relatively higher temperatures occur during a major glacial period. Such periods are called interstadials, but as yet no universally acceptable definition distinguishes an interstadial from an interglacial (Goudie, 1977).
Earlier glacial stages in Arizona probably produced climatic conditions similar to those of the last glaciation, the Wisconsin, described above. Gray ( 1961) provided paleobotanical and sedimentary evidence in the Safford Valley in southeastern Arizona to support this theory. Her evidence indicated a cooler and/or wetter climate during the early Pleistocene that she tentatively correlated with the Nebraskan glacial stage. She further suggested that winter precipitation played a more significant role in southeastern Arizona then than now. Segota ( 1967) believed that temperatures of the glacial ages were not the same, that each younger glacial age was a little colder than the preceding one. His pollen-stratigraphy data are for Europe but probably are applicable worldwide.
An interglacial was defined by Suggate ( 1965) as being a warm period between two glaciations during which the temperature rose to that of the present day. The term interglacial is used also to describe a warm episode between two cold ones in nonglaciated regions when the cold episodes correspond to those truly glacial effects that occurred at higher latitudes or altitudes (Suggate, 1974). Turner and West ( 1968) and Wright ( 1972) pointed out, however, that the climate of an interglacial was not uniform, that each interglacial consisted of distinct subperiods of vegetational development and presumed climatic conditions as evidenced by the pollen record in Europe. The last interglacial is believed to have lasted from about 128,000 to between 90,000 and 73,000 years ago (Suggate, 1974).
Some Arizona soils undoubtedly have passed through one or more interglacial. The high terraces described by Morrison ( 1965), McFadden ( 1981) and Menges and McFadden ( 1981) are thought to date back to the middle to early Pleistocene. Soils on these and similar surfaces in Arizona, therefore, have been under the climatic influences of at least two glacial and two interglacial periods.
Although moisture is not included in the definition of an interglacial, Melton ( 1965) suggested that the Yarmouth in southern Arizona was semiarid, and that the Sangamon was considerably more humid than now as well as being warm. Melton's ( 1965) Sangamon climate was postulated to explain formation of well-developed and deeply weathered red soils on old alluvial fan surfaces.
No agreed upon time that is applicable worldwide has been established for the beginning of the Holocene, time when the last glacial age ended (Mercer, 1972). Packrat midden paleobotanical evidence indicates that a rather sharp
The worldwide climate of the Holocene has changed as documented by the advances and retreats of glaciers (Denton and Karlen, 1973; Grove, 1979; Beget, 1983). In North America, the Holocene was subdivided into three phases (Bryan and Gruhn, 1964):
There are several views of the climate in Arizona during these three phases of the Holocene. Bryan ( 1941) and Antevs ( 1955, 1962) believed that the Anathermal climate had temperatures initially like those of the present, but that temperatures increased over time and the moisture levels became subhumid to humid. Martin ( 1963) believed that the climate of 10,500 to 8,000 years ago, the Anathermal, was dry as now. Van Devender and Spaulding ( 1979) indicated that the late Wisconsin winter precipitation and colder winter temperatures continued from 11,000 to 8,000 years ago. The colder temperatures were caused possibly by a decrease in the continental glaciers so that they no longer prevented cold polar air masses from entering the mid-continent. Cole ( 1982), on the other hand, suggested an early Holocene, about 9,000 years ago, increase in the summer monsoon precipitation in the eastern Grand Canyon.
Bryan ( 1941) and Antevs ( 1955, 1962) believed that the Altithermal was a long warm, dry period. Martin ( 1963) challenged the Bryan-Antevs interpretations, especially in southern Arizona where the climate of roughly 8,000 to 4,000 years ago was subpluvial, characterized by heavy summer rains. The differences between the two views lie largely in the interpretations of arroyo cutting and filling in response to climatic change and in the interpretation of pollen data. Mehringer ( 1967a) attempted to reconcile Martin's ( 1963) and Antevs' ( 1955, 1962) views by suggesting that the Altithermal ended 500 to 1,000 years earlier than Antevs proposed. Thus, according to Mehringer ( 1967a), the early Medithermal that Antevs reported as being cool-moist was actually late Altithermal. Furthermore, the pollen record in southern Arizona between Altithermal years 7,000 to 5,000 is sketchy due to limited samples (Mehringer 1967b; Haynes, 1968). Haynes ( 1968) suggested two climatic episodes within the Altithermal that may have contributed to the different interpretations of pollen data. The first part of the period was relatively dry while the latter part was characterized by more effective moisture, according to Haynes ( 1968). More recent fossil plant data from packrat middens led Van Devender ( 1977) and Van Devender and Spaulding ( 1979) to conclude that the present climatic pattern was established after about 8,000 years ago. At that time the amount of winter precipitation diminished and the summer monsoon expanded, giving Arizona its characteristic biseasonal moisture pattern. Van Devender and Spaulding ( 1979) also tended to support Martin's ( 1963) interpretation of the middle Holocene climate by suggesting that more monsoon rain fell than now.
Bryan ( 1941) and Antevs ( 1955, 1962) believed that the Medithermal in the Southwest was moderately warm with periods more moist and others more dry than now. Martin ( 1963) considered the climate of this period to resemble closely the current conditions.
Regardless of the finer points of interpreting Holocene climatic conditions in Arizona, the climate did fluctuate both in effective moisture and in temperature. And these differing climatic conditions had an effect on Arizona soils.
The mouth of Madera Canyon, Santa Rita Mountains on the Santa Rita Experimental Range. In the foreground is Sonoran desert scrub; on the alluvial fan, desert grassland; and on mountains in the background, oak-pine woodland. (Photo by Michael Parton)
*. The soil temperature refers to the mean annual soil temperature at a depth of 50 cm (20 in) or at the soil-rock interface in shallow soils.