Kuniaki Sato1, Akira Wada2, Takashi Sasaki3 and Rabindra Raj Giri1
1Saitama University, Hydroscience and Geotechnology,
255, Shimo-okubo, Urawa, Saitama, Japan
Tel: 8148-858-3570, Fax: 8148-855-1378, E-mail: sato@post.saitama-u.ac.jp
2Nihon University, College of Industrial Technology,
1-2-1, Izumi, Narashino, Chiba, Japan
Tel: 8147-474-2430, Fax: 8147-474-2449, E-mail: wada@civil.cit.nihon-u.ac.jp
3Ark Information System
Gobacho, 4-2, Chiyodaku, Tokyo Japan
Tel: 813-3234-9233, Fax: 813-3234-9402, E-mail: sasaki@ark-info-sys.co.jp
Abstract: This research is carried out with the objective of predicting quantitatively the water and heat budgets due to changes in land use through the analysis of typical surface elements, i.e., water and heat budgets in the atmosphere and soil, by the use of SALSA (Soil-Atmosphere Linking Simulation Algorithm).
Numerical solutions of momentum, heat and mass transport equations are available for prediction of physical processes occurring in atmospheric boundary layer and in porous bodies. The processes in the atmosphere and in the porous bodies are linked each other through the equations of heat and mass conservation. The water mass in the soil is treated as a two-phase mixture of liquid water and vapour. The simulated results are supported by the observation data recorded with a meteorological station placed in the experimental field in Hanno new resident town, Saitama prefecture.
SALSA has so far been applied only under simple surface conditions such as bare land. In this research, however, it is dealt with surface conditions other than bare land.
In this research, typical surface elements are divided into 4 categories, i.e., bare land, soddy land, forest and pavement, and the water and heat budgets due to these individual land uses are subjected to analysis.
The analysis of water and heat budgets by SALSA in the stage of city planning makes it possible to quantitatively predict an influence on the environment after development.
Keywords: unsaturated flow, coupled heat and water, wimulation model
With recent on-going advanced use of land in the development of new towns, large-scale housing site construction, etc., rapid urbanization is underway in large cities and their vicinities. In conjunction with such a trend, impacts on water and heat environments have begun to be feared as social problems. For example, changes in runoff systems due to development, reduction in evapotranspiration, decrease in ground water level due to reduction in rainfall infiltration, elevated urban temperatures due to increase in artificial waste heat and other effects on the soil and atmosphere due to changes in land use, have appeared. These impacts gradually change the heat energy balance or water balance in the vicinity of the ground surface and rebuild the meteorology peculiar to cities. Such effects of ground surface changes on local micro-meteorology cannot be ignored as significant social problems. To quantitatively track down these phenomena, it is important to predict variations in water and heat balances in the ground surface due to changes in land use.
Against such background, this research was carried out with the objective of predicting quantitatively the water and heat budgets due to changes in land use through the analysis of typical surface elements, i.e., water and heat balances in the atmosphere and soil, by the use of SALSA (Soil-Atmosphere Linking Simulation Algorithm).
In this research, typical surface elements were divided into 4 categories, i.e., bare land, soddy land, forest and pavement, and the water and heat budgets due to these individual uses were subjected to analysis. SALSA has so far been applied only under simple surface conditions such as bare land. In this research, however, it was decided to deal with surface conditions other than bare land.
Prior to making a numerical analysis in this research, parameters such as roughness height, permeability coefficient, etc., initial conditions and boundary conditions were established in the four surface element models, and the needed physical quantities in the atmosphere and soil were determined (soil: temperature and water content: atmosphere: wind velocity, potential temperature and moisture mixing ratio). On the other hand, in regard to input data necessary for the analysis, on-site observations were conducted on the surface-layer atmosphere and soil in the Big Hills site in Hanno City, Saitama Prefecture, for purposes of verification and predictive analysis.
The modified model in the study is composed of three main parts: surface, atmospheric and soil with equations of mass and energy conservation. For the sake of simplicity, it is assumed that the domain concerned is homogeneous in the horizon. Using measured meteorological data and surface conditions, the model can estimate the heat and mass transfer between soil and atmosphere through the soil surface. After that, the physical processes occurring in the atmospheric boundary layer and porous media are simulated. Only main methodology of the model is described in this study.
Computational algorithm in soil is linked with that in the atmospheric by the surface condition where the conservation of heat and water mass is specified. It is assumed that there is no storage of heat and mass on surface layer. The net radiation is divided into sensible, latent and ground heat fluxes, which are the condition for solving equations in both porous body and atmospheric boundary layer. The water mass transfer through the surface is composed of evaporation from soil and precipitation.
The field site where the measurement system is stationed is located in Hanno new resident town, Saitama prefecture (35o53’N, 138o36’E).
The meteorological variables: air temperatures at two heights (0.5 and 1.0m), humidity (wet- and dry-bulb), solar radiation, ground heat flux, albedo, precipitation, atmospheric pressure, wind speed and direction, have been measured at every minute. In addition to the meteorological observation, the soil samples are taken at every 5cm from surface down to 1m deep on Aug. 12, 1997 to identify the soil hydraulic properties, its geological structure and the water content with laboratory experiments. Also on this day the thermal sensors were installed to record the soil temperature at corresponding depth (See Fig.1).
Fig.l System of measurement
The objective of this analysis is to elucidate water and heat budgets between the atmosphere and the soil as typical surface elements, by use of the SALSA model. Here, calculations were made with respect to a total of 8 cases according to the existence of rainfall, for the following 4 different surface elements: 1) bare 1and, 2) soddy land, 3) forest and 4) pavement. The images of analytical cases are shown in Fig.2.
Fig.2 Images of analytical cases for surface elements
The characteristics of the surface elements can be expressed by roughness height, vegetation resistance, surface albedo, soil permeability coefficient and thermal conductivity, etc. By bare land means the ground surface condition in which no vegetation exists and the soil is exposed. As a matter of course, a minimum value is taken for the roughness height, and no vegetation resistance exists because there is no vegetation. By soddy land means the glassland whose vegetation height is about 10cm. By forest means the element model in which there are dense trees of about 10m high. In this case, the roughness is high and the vegetation resistance is large. By pavement means an image of concrete-paved road. Here, the roughness height is minimum and there is no vegetation resistance, as with the bare land. However, conditions such as thermal conductivity, thermal capacity and permeability coefficient of soil greatly vary.
As for the forest, analyses were carried out by introducing the concept of vegetation resistance in the SALSA model and reducing 80% the all-weather radiation to be input, because there are water and heat environments peculiar to the forest, such as radiation by trees, shielding by rainfall, transpiration from leaves, etc.
The simulation was carried out for a selected domain from the ground-water table up to 3,069m in the atmospheric boundary layer on the assumption that the domain is horizontally homogeneous. The groundwater table was fixed at 1.6m below the ground surface. At this boundary, the soil temperature and water content were kept constant during the entire simulation period. The porous body was divided into 25 layers by non-uniform computing meshes. The grid size was smallest (0.2cm) at the layer closest to the surface and gradually increased to 30cm in the groundwater table layer.
Fig.3 Salsa initial and boundary conditions
For the atmospheric computational domain, a total
of 11 grid points were adopted for the smallest size 3m at the surface and 1,500m
for the top layer. For the top atmospheric layer, the following assumptions were made: u = ug, v =
vg = 0,
=
= 0, H = 0, E = 0,
= 0.
If analytical conditions are observed more closely per surface element, initial conditions in the atmosphere are common to all analytical cases. However, in the case where there is rainfall, the saturated soil moisture content was input up to a depth of 0.05m from the ground surface as an initial soil condition, to express the case due to rainfall. Both rainfall intensity and rainfall duration can be expressed by varying the water content of soil and the infiltration depth of rainfall. The principal parameters used in the analyses are shown in Table 1.
When studying the analysis of water and heat balances by SALSA, a very important precondition is the addition of verification regarding whether or not the SALSA itself can be applied to the prediction of surface elements - water and heat budgets (surface element model see Fig.2). Therefore, the results of on-site measurements on the distribution of soil temperatures and water contents were compared with the results of analyses to verify the applicability of the SALSA model.
The sampling of soil and on-site measurements of soil temperature distributions at Hanno Big Hills were carried out in summer from August 6 to 7, 1998. The comparison of analytical results with measured values is shown in Fig.4. From these results, it can be confirmed that there is a tendency of agreement between the results of observations and the measured values. With similar studies made in winter as well, it was recognized that SALSA is effective to predict the temperature and water content of soil.
Fig.4 Comparison of analytical results with measured results
Among the surface elements, attention was firstly paid to the results of analyses in bare land (Fig.5) and forest (Fig.6).
At first, with regard to sensible heat, latent heat, atmospheric net radiation and soil heat fluxes on the surface, the net radiation in forest was found smaller as compared with that in bare land due to radiation shielding by trees. Also, concerning changes in the temperature of soil with the lapse of time, the soil temperature in forest was found lower as compared with that in bare land due to shielding of solar radiation by trees, decrease in wind velocity and other effects. Changes in the soil moisture content with the lapse of time were small in forest, always keeping a high water content. In bare land, the surface gets dry during the daytime, with the water content undergoing a great change. As a reason for this, the following can be said according to the results of analyses: Evaporation from the ground surface becomes small due to shielding of solar radiation in forest; in the forest soil having many voids and high permeability due to humic soil, a large amount of water can be stored in these voids; the moisture comes to reach the surface quickly due to a capillary rise from groundwater.
Fig. 5 Result of analysis in bare land (according to date obtained on July 9,1998)
Furthermore, if attention is focused on the soil temperature and water content, at the time when the soil surface temperature goes down due to a drop in atmospheric temperature during the night, the amount of evaporation from soil decreases and the water content in the neighborhood of the ground surface begins to increase. At dawn around 5 o’clock in the morning, the soil surface temperature reaches a minimum value and the soil surface water content rises to a maximum value. Also, at 1 o’clock in the afternoon during the daytime when the soil surface temperature rises to a maximum value, the soil surface water content reaches a minimum value. Little time lag is recognized between both, and it can be said that the temperature rise and the subsequent drop in the water content are in responsive relationship.
On the other hand, with regard to wind velocity (Fig.7) in the atmosphere, it is evident that at a height of 1.5m in bare land where the height of roughness is small, the wind velocity is about 5 times as much as that in the forest. There is also a general tendency that the wind velocity intensifies during the daytime when the exchange of energy on the ground surface becomes more frequent, as well as a tendency that the wind velocity calms down during the nighttime. An influence due to the height of roughness appears up to the neighborhood of 300m in altitude (upper end of a turbulent boundary layer due to surface roughness). At an altitude greater than that, however, it seems that the effect of wind velocity due to surface conditions is small.
Fig. 6 Result of analysis in forest (according to date obtained on July 9, 1998)
Next, attention was focused on the effect of rainfall on the soil moisture content, with regard to similar bare land and forest elements (Fig.8). After the rainfall, the soil in the neighborhood of a wet surface suddenly moves in the dry direction; at a point deeper than 15cm, on the other hand, the rainfall wet front moves downward from the ground surface, thereby causing the soil moisture content to rise. In the forest having a large permeability coefficient, moisturing takes place at a speed faster than that in the bare land, so changes in the soil moisture content also move toward stabilization and convergence quickly. It is clear that at a point deeper than 40cm there is little effect by rainfall infiltration from the ground surface.
Fig.7 Comparison of wind velocity distributions between bare land and forest (according to data obtained on July 9, 1998)
Fig. 8 Spatial interface elevation pattern after NW5/M/J wind
In this paper, as its objective, the possibility of predicting water and heat budgets due to surface elements in city planning was clarified by use of SALSA (Soil Atmosphere Linking Simulation Algorithm), consequently obtaining the following conclusions:
(1) It is quantitatively possible to predict water and heat budgets in four typical surface elements, by use of SALSA.
(2) The concept of vegetation resistance is incorporated in SALSA, thereby achieving some measure of success in the analysis of water and heat budgets in soddy land and forest.
[1] H.F.M. ten Berge (1990). Heat and water transfer in bare topsoil and the lower atmosphere, Centre for Agricultural Publishing and Documentation (Pudoc) Wageningen, the Netherlands.
[2] Cary J.W. (1979). Soil heat tranducers and water vapor flow, Soil Science Society of America Journal 43: pp.835-839.
[3] Fukuhara T., G.F.Pinder & K.Sato (1990). An approach to fully coupled heat and moisture transfer analysis in saturated unsaturated porous media during surface evaporation, Proc. of Japan Society for Civil Engineers (JSCE) Vol.423/II-14 pp.111-120.
[4] Milly P.C.D. (1989). Unsaturated flow induced by evaporation and transpiration, Unsaturated Flow in Hydrologic Modelling. Theory and Practice, H.J.Morel-Seytoux Eds., Kluwer Academic Publishers, pp.221-240.
[5] Philip & de Vries (1957). Moisture movement in porous materials under temperature gradients. Transactions of the American Geophysical Union 38: pp.222-231.
[6] Rawls W.J. & D.L. Brakensiek (1989). Estimation of soil water retention and hydraulic properties, Unsaturated Flow in Hydrologic Modelling. Theory and Practice, H.J. Morel-Sevtoux Eds. pp.275-300.
[7] Tennekes H. & J.L. Lumley (1972). A First Course In Turbulence. The MIT Press, Cambridge, Massachusetts.
[8] Tzimopoulos C. & E. Sidiropoulos E. (1981). Coupled heat and moisture transfer in unsaturated porous bodies, Flow and Transport in Porous Media, A. Verruijt & F.B.J. Barends, Eds., A.A. Balkema Rotterdam, pp.169-172.