{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

CharneyJAS77 - 1366 JOURNAL OF THE ATMOSPHERIC SCIENCES A...

Info icon This preview shows pages 1–20. Sign up to view the full content.

View Full Document Right Arrow Icon
Image of page 1

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
Image of page 3

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 4
Image of page 5

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 6
Image of page 7

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 8
Image of page 9

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 10
Image of page 11

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 12
Image of page 13

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 14
Image of page 15

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 16
Image of page 17

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 18
Image of page 19

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 20
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 1366 JOURNAL OF THE ATMOSPHERIC SCIENCES A Comparative Study of the Effects of Albedo Change on Drought in Semi-Arid Regions JULE CHARNEY Department of Meteorology, Massachusetts I mtitutc of Technology, Cambridge 02139 WILLIAM J. QUIRK, Snu—Hsmrv Cnow1 AND JACK KoaNrrsm2 NASA Institute for S pace Studies, New York, N. Y. 10025 (Manuscript received 31 January 1977, in revised form 17 May 1977) ABSTRACT Results from a series of numerical simulations are presented to show the effects of changes in albedo on rainfall in six areas, two each in Africa, Asia and North America. Each pair consists of a semi—arid area lying at the boundary between a major desert and an adjacent monsoonal region, and an area of the same size located within the monsoonal region itself. The sensitivity of the rainfall to the ground hydrology was determined by performing the albedo simulations with two difierent evapotranspiration parameterizations, one giving too high evaporation overland and the other giving negligible evaporation overland. In the high evaporation case, an albedo change from 0.14 to 0.35 caused large decreases of rainfall in all three of the semi-arid test areas and in two of the three monsoonal test areas. In the negligible evaporation case the simulations were performed only for the three semi-arid areas; in this case the albedo increase produced a significant decrease in rainfall in only one of the areas, the Sahel. Interoomparison of the high and negligible evaporation cases shows that changes in evaporation rate are as important as changes in albedo. Thus, in all but one of the six areas local evaporation was a major factor in influencing rainfall. When appreciable evaporation occurs, the mechanism by which an increase of albedo reduces the rainfall is as follows. Initially, the increase of albedo acts to reduce the absorption of solar radiation by the ground and therefore the transfer of sensible plus latent heat into the atmosphere. The resulting reduction in con» vective cloud tends to compensate for the increase of albedo by allowing more solar radiation to reach the ground, but it reduces the downward flux of longwave radiation even more, so that the net absorption of radiation by the ground, solar plus longwave, is decreased. Thus, with or without evaporation, the increase VOLUME 34 of albedo causes a net decrease of radiative flux into the ground and therefore a net decrease of convective cloud and precipitation. ‘ 1. Introduction It is widely feared that man is inadvertently altering the delicate balance of the present climate (SMIC, 1971). This balance is especially precarious in the semi-arid zones bordering on the major deserts. Re— cently severe drought occurred in the Sahel at the southern margin of the Sahara when moist monsoon air failed to penetrate sufliciently far northward from the tropical Atlantic. Drought has been a recurrent problem in the provinces of India and Pakistan which border on the central Asiatic Deserts. In North America the highly productive lands of the western Great Plains lying just east of the Great Western Desert have suffered extended periods of drought, including the "dust-bowl” drought of the 1930’s and 1Present afliliation: Computer Sciences Corporation, Silver Spring, Md. 20910. ’Present afliliation: Volcani Institute, P03 6, Bet Dagan, Israel. the Widespread drought of the 1950’s. These regions have in common that large changes repeatedly take place in plant cover and ground hydrology and that maximum rainfall occurs during the summer monsoon season and is primarily convective in character. In his Royal Meteorological Society Symons Me- morial Lecture for 1974, Chamey (1975) showed that . the radiative heat loss caused by the high albedo of a desert contributes significantly to the sinking and drying of the air aloft and therefore to the reduction of precipitation. This dependence of precipitation on albedo led him to propose a biogeophysical feedback mechanism linking vegetation, albedo and precipita- tion as a partial explanation for recurrent drought in desert border areas. He suggested that if the soil is light, dry and sandy, as it often is in these areas, a decrease of vegetation will lead to an increase of albedo, a reduction of precipitation and therefore a. further decrease in vegetation or at least a per- petuation of the initial decrease. This efiect was SEPTEMBER 197 7 expected to be further enhanced by the reduction in evapotranspiration resulting from the decrease in vegetation. In his view the vegetative cover at the border of a desert is to be regarded as unstable or metastable; if unstable, an initial departure from equilibrium will amplify; if metastable, it will at least tend to persist. The perturbation will remain until external counteracting climatological forces become large enough to overcome the local feedback effect and return the system toward the old equilibrium. These forces act to prevent still another possibility, namely, that if they did not exist or were sufficiently weak, the positive feedback could lead eventually to a new statistical equilibrium, i.e., to a new climate. Such a system would be intransitive in the sense defined by Lorenz (1968), having the property that difierent initial states lead to different climates. But if the external forces are sufliciently strong, the new statistical configuration cannot persist indefinitely, it can exist merely as a long-term perturbation of a basically stable statistical equilibrium. Albedos can be changed by natural as well as anthropogenic factors. We do not distinguish between the two here, but if the entire system were being considered, in its biological and ecological as well as its geophysical aspects, a sharp distinction would have to be made, since anthropogenic changes involving changes in land use tend to be very persistent whereas natural ecosystems tend to recover more readily from natural perturbations. In this article we present modeling evidence only for the reality of the geophysical, not the biological, component of the system. We show that the restoring climatological forces, consisting of the global circula- tions and their random perturbations, are too weak for the limited period of the integration to overcome or even mask the local effects of the albedo changes in the regions in question. Hence, whatever the ulti- mate behavior of the biogeophysical system as a whole, the feedback mechanism will exist and will have the effect of at least prolonging a given fluctuation, whether positive or negative, i.e., of prolonging both dry and rainy periods. Since drought is prolonged dryness, the postulated feedback mechanism can be expected to contribute to drought. Control numerical integrations were carried out for the month of July with the general circulation model of the NASA Institute for Space Studies. These were repeated with surface albedos modified in certain selected regions: the Sahel, the western Great Plains and the semi—arid Rajputana area of southeastern Pakistan and northwestern India adjoining the very dry Thar or Indian Desert. For comparison purposes, a third integration was performed in which the albedos were altered simultaneously in central Africa, the Mississippi Valley and Bangladesh. These regions were selected as being rainier and more centrally imbedded CHARNEY, QUIRK, CHOW AND KORNFIELD 1367 in their respective monsoon regimes. Some early results for the Sahel integrations had already been presented by Charney el al. (1975, 1976). Man~induced changes may very well have operated in the semi—arid zones dealt with in this article. In the western Great Plains, for example, erosion, agricultural tillage and crop harvesting have at times reduced the darker humus layer to expose the brighter lower horizons of the soil (Buckman and Brady, 1969). In the Sahel, there is strong evidence that overgrazing has caused surface albedos to increase over large areas (Otterman et al., 1976) and as reported by Hora (1952), “the Rajputana Desert is largely a man—made desert . . . (formed) by the work of man in cutting down and burning forests . . . (and by) the deterio— ration of soils.” The last reference is taken from an article by Bryson and Baerreis (1967) who suggest that the dense pall of dust frequently found over northwestern India is a significant factor in increasing the infrared cooling of the air and thereby increasing its subsiding motion. Since destruction of vegetation destabilizes the soil and permits the wind to raise more dust, the pos- sibility exists that the dust enhances our postulated biogeophysical feedback mechanism. However, dust also absorbs solar radiation, and it is not yet known whether it produces a net increase or decrease of radiative heating. Some early estimates suggest that the dust does not greatly change the planetary albedo but that it absorbs solar radiation in the mid-atmo- sphere and thereby reduces the heating of the ground (J. Hansen, personal communication, 1976). This effect would tend to stabilize the lower part of the atmosphere and inhibit thermal convection. It is not considered here. Since accurate measurements of actual albedo changes on the earth have not yet been made, we have had to content ourselves with idealized numerical experiments with simplified albedo distributions and arbitrary albedo changes. To the extent that the albedo~induced perturbations are linear, their mag- nitudes can later be scaled for observed changes in albedo as these become available. The albedos for all permanent desert areas in the Northern Hemisphere were fixed at 0.35, and the albedos of the marginal and comparison areas were assigned the values 0.35 or 0.14; the albedos of all other land surfaces save permanent ice cover were taken to be 0.14; the ice albedo was assigned the value of 0.70, and the ocean surface albedo was taken to be 0.07. The albedos selected turn out to be within the range determined by Rockwood and Cox (1976) from satellite observa- tions over northwestern Africa. The locations of the permanent desert and marginal areas were assigned on the basis of relative regional brightness obtained from minimal planetary albedos (Rashke et al., 1973). To insure computational sig— ' v-u .— 40:“; 1 , -/ . 6° :0 "‘y vvv . ’0 .93.. v .0 O . .0‘ 40 v . . . 9:0}: 0 o o 8 / m 0 2° IDEAUZED PERMANENT DESERT IDEALIZED MARGINAL AREAS 4O HUMID AREAS ”ICE COVER 80 WW] / // 4 / >0 0 1 , WI$W// 180 ISO :40 120 100 so 6040 2. JOURNAL OF THE ATMOSPHERIC SCIENCES ./ VOLUME 34 (a (Illl' 180 ISO 60 BO IOO |20 I40 FIG. 1.1. Global albedo distribution assumed for experiments. nificance the western Great Plains and Rajputana areas were made somewhat larger than they are in reality. Fig. 1.1 shows the albedo distributions as— sumed for the experiments. 2. The model The numerical general circulation model employed for the studies was that of the Goddard Institute: for Space Studies (GISS). It has been described by Somerville et al. (1974), and its July climatology for .a land albedo of 0.14 has been discussed by Stone et al. (1977). The model contains two mechanisms for generating cloud and precipitation: 1) large-scale ascent leading to supersaturation, and 2) cumulus convection in conditionally unstable air. The cumulus convection parameterization is the same as that de- vised by Arakawa for the UCLA three-level model (Arakawa et al., 1969). In the GISS nine-level model used for the experiments, the lower six levels were combined in pairs to produce three layers for the calculation of cumulus convection. Model convection occurred whenever a lower layer was buoyant with respect to a higher level. Convection stabilized the column, condensed water vapor, released latent heat and created clouds for input into the radiation portion of the model. Clouds arising from both types of con- densation mechanisms were assumed to occupy an entire layer and to be black in the infrared. The effect of a cloud on solar radiation was made to de- pend on the cloud type, cloud depth and cloud height. The hydrological parameterizations used in the model were exceedingly primitive. The ground wet— ness G, defined as the fraction of saturation of the soil, was fixed at values determined from mean July surface relative humidities (Stone et (11., 1977). The evaporation rate E was taken to be a fraction ,8 of the potential evaporation rate P, the evaporation rate from saturated ground. In one set of integrations the coeflficient B was set equal to the lesser of 1 or ZG, as in the Arakawa model, but this assumption led to excessive evaporation over land. In a second set the empirical formula 1 — exp[44.6G/ (1 —0.99G)] exp(0.87P) was introduced as being more typical of plant physio— SEPTEMBER 1977 CHARNEY, QUIRK, CHOW AND KORNFIELD 1369 TABLE 3.1. Detailed description of six model runs discussed in the text. Numerical simulations Albedo Experiment Evaporation no. Land Ocean Ice and snow Desert* Marginal“ Humid*** over land 1 0.14 0.07 0.7 0.14 0.14 0.14 Excessive 2a 0.14 0.07 0.7 0.35 0.14 0.14 Excessive 2b 0.14 0.07 0.7 0.35 0.14 0.14 Negligible 3a 0.14 0.07 0. 7 0.35 0.35 0.14 Excessive 3b 0.14 0.07 0.7 0.35 0.35 0.14 Negligible 4 0.14 0.07 0.7 0.35 0.14 0.35 Excessive * Desert: Sahara, Nevada, Sonora, Gobi, etc. (see Fig. 1.1). ** Marginal: Sahel, Western Great Plains, Rajputana (see Fig. 1.1). a" Humid: Bangladesh, Central Africa, Mississippi Valley (see Fig. 1.1). logical characteristics. Unfortunately, the model con- siderably overestimated P and thus greatly reduced the values of E, so much so that the evaporation over land became nearly zero. Pending the completion of a more realistic model of evapotranspiration, which takes into account the variable storage of moisture in the ground and the ability of plants to enhance evaporation by drawing upon deeper lying moisture, it has been decided to present the results from the two primitive hydrologies as at least bracketing the effect of evaporation over land. The calculations with the excessive evaporation hydrology will be referred to as “excessive” or “high” evaporation experiments and those with the negligible evaporation hydrology as “negligible” or “low” evaporation experiments. 3. Model climatology for July The actual state of the atmosphere on 18 June 1973, determined from data supplied by the National Meteorological Center, was taken as the initial state for the global integrations. The integrations were carried out for the six-week period (18 June—31 July). The sea—surface temperature and ice cover, as well as soil moisture, were fixed at their mean climatological values for July throughout the integration. The solar declination, solar insolation and snow lines were al- lowed to vary with time. The distributions of soil moisture, sea—surface temperature and ice cover are described in Stone et al. (197 7). The parameters used in the various numerical experiments and their geo- physical distributions are listed in Tables 3.1 and 3.2. By model “climatology” for July we mean merely an average for the month of July. The “climatology” described by Stone et al. corresponds to the first ex- periment in Table 3.1, in which the albedo over ice— free land was everywhere 0.14 and the excessive evaporation model hydrology was employed. The assumption made here and throughout this article is that the calculated mean conditions for July are in fact a reasonable approximation to the model climatology for July. Thus when the integrations were extended two weeks into August for Experiments 1 and 2a, no significant changes were observed in the mean conditions. But a true statistical equilibrium would require numerical calculations for many model years. And it would not suflice to keep the sun fixed in the heavens because the atmosphere-earth system has relaxation times which are long compared to a month, as for example for water storage in the~ ground or heat storage in the upper mixed layer of the oceans. a. Influence of albedo on rainfall and evaporation in the Northern Hemisphere The simulated precipitation rates for the first and third experiments are compared with observations in Figs. 3.1. A comparison of Figs. 3.1a and 3.1b shows that while the pattern of rainfall agrees qualitatively with observations for the integrations with the ex— cessive evaporation hydrology and low albedos over deserts and semi-arid regions (Experiment 1), the rainfall in the Northern Hemisphere deserts is about twice the observed. Comparison of Figs. 3.1b and 3.1c L (Experiment 3a) shows that the increase of desert albedo and desert margin albedo from 0.14 to 0.35 results in more realistic precipitation rates over the deserts. This is seen in Table 3.3 which compares the calculated with the observed precipitation rates in the various experimental regions shown in Fig. 1.1. Comparison of Figs. 3.1b—3.1d (Experiment 3b) shows TABLE 3.2. The latitudinal and longitudinal boundaries of the various test regions. Number Geographical of grid region points Latitude Longitude Marginal Sahel 11 16°N—20°N 17.5°W—37.5°E Rajputana 4 24°N—32°N 67.5 °E—77.S°E Western Great Plains 8 32 "N—48°N 107.5 °W—97.5 °W Humid Central Africa. 13 8°N—12°N 12.5 °W—52.5 °E Bangladesh 4 20°N-28 °N . 7 7.5 °E—87.5 °E Mississippi Valley* 7 32°N—48 “N 92.5°W—82 .5 °W * Except 44—48”N and 82.5—87.5“W representing the Great Lakes. 1370 JOURNAL OF THE ATMOSPHERIC SCIENCES l20 I40 IOO l20 I40 FIG. 3.1 (3.). Observed June—August mean rainfall (mm day“) [Moller (1951) as analyzed by Schutz and Gates (1972)] and July mean rainfall (mm day—1) for (b) the excessive evaporation model with desert and desert margin albedo=0.14, Experiment 1; (c) the excessive evaporation model with desert albedo and desert margin albedo=0.35, Experiment 3a; and (d) the negligible evaporation model with desert albedo and desert margin albedo=0.35, Experiment 3b. VOLUME 34 SEPTEMBER 1977 CHARNEY, QUIRK, CHOW AND KORNFIELD 1371 TABLE 3.3. Calculated and observed precipitation rates (mm day—1) over desert and monsoon regions in July. Desert regions Sahara and Arabian Middle Eastern Desert Central and Eastern Asiatic Desert Great Western Desert Monsoon regions Central Africa Mississippi Valley Calculated Albedo = 0.14* Albedo = 0.35““ Observed 4.21 2.63 0.18 4.75 1.09 0.46 3.40 1.96 1.24 2.37 1.01 0.47 4.16 6.27 5.07 3.54 2.23 1.94 9.31 7.64 7.94 Bangladesh * Experiment 1 in Table 3.1. ** Experiment 2a in Table 3.1. that the decrease in precipitation due to the sup— pression of evapotranspiration is as large as the de- crease due to the increase of albedo. In both high desert albedo cases the rainfall maximum in the Sahara remains some 4° latitude too far north. The remaining figures described in this section pertain to experiments with high albedos for both desert and semiarid regions (Experiments 3a and 3b). Zonal averages of precipitation were computed for land and ocean for both the excessive and negligible evaporation hydrologies. These are shown in Fig. 3.2. Over land, the rainfall is too high in the high evapora- tion case and too low in the negligible evaporation case. Also, the tropical rainbelts of both models are Excessive Evaporation Neql i 9i bie Evaporation Observation Precipitation (mm/day) 4020 0 20 Latitude 40 60 BON about 4° latitude too far north (Fig. 3.2a). Over the oceans, both hydrologies give too little extratropical rainfall, and the computed maximum tropical rainfall is at the equator whereas the observed maximum is nearly 10° north of the equator (Fig. 3.2b). The Northern Hemisphere distributions of observed and simulated evaporation rates are compared in Figs. 3.3a and 3.3b for the excessive evaporation hydrology.3 The model evaporation rates are much larger than the observed over land and much lower over the oceans. Longitudinal averages over land are shown in 3What are called “observed” rates are actually inferences from the surface heat balance. Excessive Evaporation ........... Negligible Evaporation —-- —— ...
View Full Document

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern