Climate changes during the Upper Pleistocene and Holocene

Davis (1995) has built a database of more than 1000 samples from over 50 sources. By applying the analogue technique, he reconstructed the paleo-climates of three sites: the Montezuma Well in Arizona, Exchequer Meadows in California and Rattlesnake Cave in Idaho.

Examining the Montezuma Well paleo-temperature curve, one can divide the Holocene sequence into three stages.

1. The early cold stage, 10 ka to 8 ka BP, in which the range of temperatures was below 12 °C level (c. 10 °C).

2. The mid-warmer stage, 8ka to c. 2kaBP, when the range of temperatures was above 12 °C with a few peaks approaching 15 °C.

3. The upper colder stage, from 2 ka BP to present, with temperature fluctuations around 12 °C, with four minimum levels, of which two (at 1.6 kaBP and one in the twentieth century) were just below the 10 °C level.

Comparing the temperature curve with the precipitation curve at the same site, it can be seen that during stage 1, precipitation was rather high, exceeding 400 mm, and reached extremes of more than 600 and even 700 mm during short periods. By comparison, stage 2 was a warmer period and precipitation ranged between 300 and 400 mm, except for a short period around 7 ka BP, when it approached the 500 mm level. During the upper stage 3, the range was more or less similar, although it has had a trend of increasing precipitation. From this, one would conclude lower temperatures corresponded with higher precipitation levels in the Holocene in Arizona.

Evidence from Exchequer Meadow in California indicated that changes in temperatures during the Holocene were less pronounced here than in Arizona. Temperatures did not go below 8 ° C and did not exceed 9 °C during the whole of the Holocene, except during stage 1, which was a little cooler. During Stage 2, temperatures reached 9 °C, while Stage 3 was somewhat cooler. Precipitation at the beginning of this stage was rather low, c. 600 mm, but has increased more or less constantly up to the present, reaching an average of about 1200 mm in recent decades. From these data, it is not possible to show a direct relationship between temperature and climate in California for the whole Holocene.

Fig. 5.1. Map of southwestern USA.

The pollen curve from Rattlesnake Cave, Idaho showed a rather warm phase during stage 1, declining towards 5kaBP and then declining again c. 2kaBP (i.e., during stage 3). Levels of precipitation showed a marked correspondence of low temperatures and higher precipitation, and vice versa.

From the palynological time series Davis (1995) could discern three stages of climate changes for the Holocene in the western USA. In the southern region, both temperatures and temperature impact on precipitation were very variable, while there was greater consistency in the north. This may be because the climate in the south was influenced by more factors than that in the north.

Davis (1992) investigated the pollen assemblages deposited in the San Joaquin marsh, situated near the sea in southern California. The pollen record traces the changes in the sea levels, as the marsh became salty during high sea levels and halophytic taxa dominated. When the sea retreated, freshwater-type vegetation dominated. Davis (1992) correlated the sea retreat episodes with periods of global cooling. These occurred, according to this reconstruction, at 3.8ka, 2.8ka, 2.3ka and after 0.56kaBP.

Hughes and Graumlich (1996) used the extensive dendro-climatic data existing for the western USA to reconstruct the precipitation for the Great Basin (Fig. 5.1) from c. 8 kaBP (Fig. 5.2). This reconstruction showed that the annual precipitation was rather uniform, ranging around 200 mm. Despite the generally uniform nature of the precipitation record, droughts and periods of above average precipitation lasting several centuries still occurred. Prolonged droughts were more persistent during the middle of the first millennium BC, as well as in the sixth and eighth centuries AD. Generally speaking, the conclusions of Hughes and

Graumlich (1996) are in general agreement with the conclusions of Davis (1992), with regard to the precipitation for the Montezuma Well area in Arizona. However, according to Davis, the precipitation from 8 ka to 6 ka BP was somewhat higher.

Graumlich and Lloyd (1996) also reconstructed the paleo-climatic fluctuations since c. 3.5 kaBP for the Sierra Nevada based on dendrochronological data. They observed periods of extended droughts during 800-59, 1020-70, 1197-1217, 12491365,1443-79,1566-1602,1764-94,1806-61 and 1910-34 AD. On the basis of records of sub-alpine tree growth and density, they concluded that between 1450 and 1850 AD temperatures were below the long-term average. Yet, as a general conclusion, they claimed that "the dendroclimatic records, if interpreted without reference to other proxy data sources, imply that climate in the late Holocene has been stationary".

Leavitt (1994) observed a strong depletion in 513C in Bristlecone Pine tree rings in the White Mountains, from 1080 to 1129 AD, which would suggest a period of abundant soil moisture allowing the stomata of the tree to remain open. In general terms, the series of drought years identified by Graumlich and Lloyd (1996) on the basis of dendrochronological data is in good agreement with Leavitt's findings based on 513C data.

Lloyd and Graumlich (1997) found that the tree line in the Sierra Nevada was at a low altitude from c. 1100 to 600 BC; it was low for a relatively short period c. 100 BC, declined sharply c. 1100 AD and remained at that level from c. 1500 to 1600 AD. One can assume that these declines were a result of cooler periods and vice versa.

Comparing the precipitation curve of Graumlich and Lloyd (1996) with the tree-line (temperature) curve of Lloyd and Graumlich (1997), and taking into account the fact that their

Fig. 5.2. Paleo-hydrology of southwestern USA.

dating is rather accurate, one can conclude that changes in temperature correlate differently with changes in precipitation in different periods: during certain periods with a high-altitude tree line (i.e., periods of warmer temperatures), precipitations were higher, while during other periods, the opposite situation occurred (when there were lower tree lines, indicating lower temperatures, there was also higher precipitation), especially during the last millennium.

Comparing Hughes and Graumlich's (1996) reconstructed precipitation curve with the levels of Lake Mono (Stine, 1994; Fig. 5.2), a rather good correlation can be found between periods of low precipitation and ones of low lake levels (assuming Stine's dates have a 100 year inaccuracy range). This is also consistent with the conclusions arrived at by Graumlich (1993).

Somewhat similar results are to be obtained from the work of Fritts and Shao (1992). They reconstructed the climate for the western USA starting in 1600 AD, based on calibration with data from the period for which instrumental data are available, namely from 1918 to 1961. Altogether, the reconstruction showed low-frequency variations, including the period of the Little Ice Age. Low temperatures and higher precipitation in the northern parts of the western USA and warmer and dryer conditions to the south characterized the period from 1602 to 1636. Around 1667, the reconstruction of climate showed a uniform decline over the western USA, except for the California-Dakota area, where precipitation seems to have increased because of strengthening of rainstorms over the California-Dakota storm track. Overall, the reconstructed paleo-climates for the last 400 years show considerable variability in all factors concerning differences in the trends of warming and cooling, as well as in temperature impact on precipitation, in the different parts of the western USA. This indicates that during this period, which was characterized by warmer temperatures over most of the western USA (except the Great Basin), there was an increase in precipitation in the southwest, diminishing in the more northern part of the western USA. This trend was connected to a period of low pressures at sea level over the entire North Pacific and especially over the North American Arctic.

The reconstruction of the climate of southwest USA by Davis (1996), was based on percentages of aquatic types of pollen. He emphasized two warm periods: the Medieval Warm Period, c. 1000-1200 AD, and the early Holocene, c. 8ka to 10kaBP. During these periods, the monsoonal summer precipitation in the southern part of the American southwest was higher, while in the northern area, it was lower. The more humid climate had a positive impact on the socio-economy of the agricultural native American societies of Arizona that started to grow c. 500 AD, reaching a climax 1100-1200 AD.

Petersen (1994) found a similar positive influence of a warm climate from c. 900 to 1300 AD on the socio-economic system of the Anasazi-Pueblo Native American people in the southern Rocky Mountains. He based his conclusions on the correlation between the archaeological data portraying the expansion of the farming societies and proxy-data, related to the movement of the lower spruce forest border in one site and the timber line at a second site. At the first site, Beef Pastures, the data were obtained by comparing ratios of spruce pollen to pine pollen, while at the second site, Twin Lakes, the data were based on the ratios of conifer pollen to non-arboreal pollen.

Davis (1994) also showed the increase in summer precipitation, caused by strengthening of monsoonal rains during warm periods, by comparing palynological indicators of lake levels with Petersen's (1994) conclusions. He suggested a correlation between these events and increased solar activity.

Scuderi (1990), who investigated tree ring variations at Cirque Peak in the southern part of the Sierra Nevada in California, found a marked correspondence, on the decadal level, between tree ring variations from the temperature-sensitive upper timberline sites in the Sierra Nevada and the sulfur-rich aerosols recorded in the Greenland ice cores. The rate of decrease of temperature was in the order of 1 °C for up to 2 years. He suggested that clusters of volcanic events may have served as triggers of climate changes but there may have been significant spatial variations in the actual changes occasioned by these eruptions. While some regions may have become cooler, others became warmer. He showed that periods when extremely narrow tree rings were produced in the temperature-sensitive sub-Alpine tree Pinus balfuriana were related to major volcanic eruptions, recorded historically or in ice cores. These periods also coincided with periods of the advance of glaciers in the Sierra Nevada. The most pronounced periods were 500 BC to 50 AD, 150-250, 540-640, 800-950 and 1600-1890 AD. Intervals of decreased volcanic activity coincided with few or no dated glacial moraines and a minimal number of wider rings; these periods occurred in 950-500 BC and 50-150, 300-450, 600-750, 1000-1250 (Medieval climatic optimum) and 1890-1980 AD. Scuderi (1993) reconstructed the mean June to January temperature at the timberline site for the last 2000 years based on the tree rings. He observed a 125 year periodicity, which may be linked with solar activity. The changes that interfered with this periodicity he attributed to chaotic solar behavior.

Pisias (1978, 1979) reconstructed a time series curve of paleo-temperatures based on the analysis of Radiolaria assemblages in an annually laminated sediment core. Using this time series, a dataset reconstructed from all observed sea-surface temperatures and hydrographic data for the California Current, he developed a model to reconstruct sea-surface temperatures and dynamic height anomaly distributions for the California Current during the last 8000 years. This reconstruction indicated that the flow of the California Current was much stronger when sea-surface temperatures were low than it was during warm periods, like that of the present.

McKenzie and Eberli (1987) investigated the oxygen isotope stratigraphy of the sediments of the Great Salt Lake, Utah (Fig. 5.2). They found a good correlation between heavier values of the isotopes and the presence of aragonite. This could be correlated with periods of decreased freshwater inflow to the lake. Their reconstruction of the history of the Great Salt Lake correlated well with the history of paleo-environments of the northeastern Great Basin during the Holocene reconstructed by Currey and James (1982). According to this latter reconstruction, during the period from 7 ka to 5.5 ka BP, the Great Salt Lake was at a very low level, almost near a stage of complete desiccation. Wetter periods around 5ka, 3.5ka and 0.6kaBP followed. These were cold periods of increased westerly flow and winter precipitation. In between, were warmer periods of low lake levels, associated with a strengthening of the Mexican monsoon activity. This happened at c. 4.5 ka and 1.5 kaBP.

By analyzing glacial deposits, Porter (1981) summarized the fluctuations of Holocene glaciers in western North America. In the curve he composed, one can observe four peaks of glacial advance before that of the Little Ice Age. These were c. 5 ka, 3 ka and 1kaBP.

From the paleo-hydrological synthesis of Baker et al. (1995) it can be concluded that dry conditions existed during the Middle Holocene over the deserts of North America.

Pollen analysis from Wildcat Lake, Whitman County, Washington for the last 1000 years (Davis et al., 1977) showed that terrestrial pollen percentages were relatively stable prior to the introduction of horses, sheep and cattle, but that changes occurred in the Wildcat Lake aquatic environment following deposition of volcanic ash from Mt. St. Helens (0.4ka to 0.5 kaBP). Two periods of intense erosion followed the introduction of grazing and the ensuing range deterioration. It is obvious from these studies at Wildcat Lake that no climatic event of the past 1000 years resulted in vegetational changes as great as those brought about by European agricultural and grazing practices.

Stine (1998) focused on the period from c. 850 to 1325 AD. The first peak of cool and dry climate was during the first century since many lakes in western America dried up c. 850 AD. Stine suggested that this cold and dry spell was on a global scale as it correlated well with the advance of some Alaskan and Canadian glaciers. Another peak of cool and dry climate occurred for c. 50 years around 1130 AD. The general cold and dry conditions changed after 1325 AD, when the water levels of the desiccated water bodies started to rise again.

Ely et al. (1993) investigated the history of paleo-floods in Arizona and southern Utah. Their study was based on the mapping of fine deposits in backwater zones, preserved in protected niches outside the minimum flow bed of the river. The mapping of these deposits, along 19 riverbeds, enabled 251 extreme flood events to be distinguished, mainly clustered in the periods 5.2ka to 3.6 ka, 2.2ka to 0.8 ka and 0.6 kaBP to the present. From 3.6 ka to 2.2 ka BP, no sediments of large floods were recorded. A fall in the number of large floods occurred between 0.8 and 0.6 ka BP, immediately following a period of frequent large floods from 1 ka to 0.8kaBP. According to these authors, the extended periods of large floods coincided with periods of cold climate, glacial advance and vegetation changes. By correlating the paleo-flood data with modern flood events and synoptic conditions, the authors found that large floods were connected to storms associated either with north Pacific winter fronts or with late summer and fall Pacific tropical cyclones, as well as with local convective summer storms. The two first conditions most probably caused (and would continue to cause) the largest floods. There was also a concurrence between large floods and El Nino events.

Correlating the climate changes of southwestern USA with those of the Levant and their impact on the hydrological cycle, one can see that during the warm and dry period of c. 4kaBP in the Levant there is an increase in precipitation in southwest USA (Hughes and Graumlich, 1996). The same warming and increase of precipitation can be concluded from the data from Montezuma Well (Davis, 1995). The isotopic data from Great Salt Lake (McKenzie and Eberli, 1987) also indicated a warming trend.

The other isochrone most probably of global significance is that of the cold period starting c. 3.5kaBP (Iron Age in the Levant, Jomon cold stage in Japan). This climatic trend was obvious in the two time series from the coast of California (Davis, 1995; Pisias, 1978). In Arizona, a general trend of less precipitation was observed (Hugues and Graumlich, 1996) and in the Great Salt Lake, one can see the starting of a trend towards depletion of oxygen isotopes (McKenzie and Eberli, 1987).

The cold period starting c. 2.3 ka BP and extending to 1.4 ka BP, with a short warm period from 1.7ka to 1.6kaBP (the Roman-Byzantine in the Levant and the Kofun period in Japan), can be observed in the two time series curves along the coast of California (Davis, 1992; Pisias, 1978). Lake Mono in eastern California was at a low level during this period (Stine, 1994). Analysis of pa-lynological assemblage at Montezuma Well in Arizona (Davis, 1995) indicates lower temperatures in the lower altitude of the tree lines during this period. The tree line in the Lake Wright Basin (Lloyd and Graumlich, 1997) was also rather low while the ratio of spruce to pine in the southern Rocky Mountains (Petersen, 1994) was rather high. This coincided with a period of reduced precipitation on the Great Basin (Hughes and Graumlich, 1996). The 18O isotope composition of the sediments of the Great Salt

Lake was highly depleted, which implies minimum evaporation and thus a high lake level, and most probably a depleted composition of the precipitation to start with (McKenzie and Eberli, 1987).

Stine (1998) discussed at length the impact of the warm period from c. 1.3 ka to 1 ka BP and termed it the Medieval Climatic Anomaly (Arab period in the Mediterranean, the Nara-Heian warm stage in Japan, the Medieval Warm Period in Europe). The temperature of the sea surface off eastern California (Pisias, 1978) was mostly high. In the Great Salt Lake data, most of this period is characterized by a heavy isotope composition, indicating, in particular, high evaporation.

Comparing all these changes with variations in the width of the tree rings of Bristol Cone Pine growing near the upper tree line on the White Mountains in California (LaMarche, 1974) raises some doubt with regard to the conclusion of LaMarche that the width is an indication of variations of summer warmth and/or its seasonal duration. One gets a better correlation between the width of the tree rings and colder periods, most probably because of higher precipitation.

Some general conclusions can be drawn with regard to climatic changes and their impact on the hydrological cycle in the western USA during the Holocene. During cold periods, there was a southern shift of the westerlies zone, resulting in an abundance of winter rains and floods in the regions affected by this system. During warm periods, there was a shift northward of the summer rain monsoonal system and summer storms may have prevailed. However, these did not cause large floods, except during periods of El Nine) events. This may be because the Colorado basin lies on the edge of the summer rain region and, therefore, storms that come from the Caribbean or Pacific tend to be moderate.

As for the past being a key to predict the future, one should also consider that the Colorado basin is on the border of two climate systems and is not far from the Pacific coast. Consequently, it is rather strongly influenced by El Nino events, making the prediction of its climate regime, and thus its hydrology, rather difficult. Although it can be generally foreseen that global warming would result in the summer storm belt moving northward, one cannot predict the extent of this shift. Overall, it can be said that a warm global change would cause higher summer precipitation rates and floods in the southern part of the Colorado basin, while in the northern part, the winter and summer rates of precipitation would be less. This scenario may change from time to time during El Nino events and/or volcanic events of large magnitude, which affect the transparency of the atmosphere on a global scale.

For California, it can be said that, in general, this region will see more summer rains coming from the Pacific, but winters will be drier. Again, during periods of El Nino events, torrential winter storms can be forecast.

0 0

Post a comment