• Bureau of Environment, cited. 2007: The environment of Tokyo 2006. Tokyo Metropolitan Government Environmental White Paper 2006. [Available online at https://www2.kankyo.metro.tokyo.jp/kouhou/env/eng_2006/index.html.].

  • Chandler, T. J., 1965: : The Climate of London. Hutchinson, 292 pp.

  • Chongqing Municipal Bureau of Statistics, cited. 2007: Chongqing Statistical Yearbook 2006. [Available online at http://www.cqtj.gov.cn/tjnj/2006/yearbook/indexe.htm.].

  • Cotton, W. R., and Coauthors, 2003: RAMS 2001: Current status and future directions. Meteor. Atmos. Phys., 82 , 529.

  • Dyer, A. J., , and B. B. Hicks, 1970: Flux gradient relationship in the constant flux layer. Quart. J. Roy. Meteor. Soc., 96 , 715721.

  • Environmental Protection Agency, cited. 2005: Global warming—Climate. Web page of the U.S. Environmental Protection Agency. [Available online at http://yosemite.epa.gov/oar/globalwarming.nsf/content/climate.html.].

  • Hadano, K., , T. Izumi, , D. Nakayama, , and H. Matsuyama, 2004: The spatial representativity of temperature data of AMeDAS as revealed by the variogram. (in Japanese). Theor. Appl. GIS, 12 , 3546.

    • Search Google Scholar
    • Export Citation
  • Ichinose, T., , K. Shimodozono, , and K. Hanaki, 1999: Impact of anthropogenic heat on urban climate in Tokyo. Atmos. Environ., 33 , 38973909.

    • Search Google Scholar
    • Export Citation
  • Kondo, H., 2001: Simulation in the field of atmospheric environment. (in Japanese). J. Comput. Fluid Dyn., 9 , 133140.

  • Kondo, H., , Y. Kikegawa, , Y. Genchi, , and S. Yamamoto, 1999: Heating in the urban canopy by anthropogenic heat use. Proc. 15th Int. Congress of Biometeorology and International Conf. on Urban Climatology, Sydney, NSW, Australia, International Society of Biometeorology (ISB) and World Meteorological Organization (WMO), CD-ROM, 249.

  • Kondo, J., , O. Kanechika, , and N. Yasuda, 1978: Heat and momentum transfers under strong stability in the atmospheric surface layer. J. Atmos. Sci., 35 , 10121021.

    • Search Google Scholar
    • Export Citation
  • Kusaka, H., , and F. Kimura, 2004a: Coupling a single-layer urban canopy model with a simple atmospheric model: Impact on urban heat island simulation for an idealized case. J. Meteor. Soc. Japan, 82 , 6780.

    • Search Google Scholar
    • Export Citation
  • Kusaka, H., , and F. Kimura, 2004b: Thermal effects of urban canyon structure on the nocturnal heat island: Numerical experiment using a mesoscale model coupled with an urban canopy model. J. Appl. Meteor., 43 , 18991910.

    • Search Google Scholar
    • Export Citation
  • Kusaka, H., , H. Kondo, , Y. Kikegawa, , and F. Kimura, 2001: A simple single-layer urban canopy model for atmospheric models: Comparison with multi-layer and slab models. Bound.-Layer Meteor., 101 , 329358.

    • Search Google Scholar
    • Export Citation
  • Landsberg, H. E., 1981: The Urban Climate. Vol. 28. Academic Press, 275 pp.

  • Martilli, A., , A. Clappier, , and M. W. Rotach, 2002: An urban surface exchange parameterization for mesoscale models. Bound.-Layer Meteor., 104 , 261304.

    • Search Google Scholar
    • Export Citation
  • Masson, V., 2000: A physical-based scheme for the urban energy budget in atmospheric models. Bound.-Layer Meteor., 94 , 357397.

  • Ministry of the Environment, cited. 2004: Outline of the policy framework to reduce urban heat island effects. Inter-Ministry Coordination Committee to Mitigate Urban Heat Island, Ministry of the Environment, Japan. [Available online at http://www.env.go.jp/en/air/heat/heatisland.pdf.].

  • Orlanski, I., 1975: A rational subdivision of scales for atmospheric processes. Bull. Amer. Meteor. Soc., 56 , 527530.

  • Pielke, R. A., and Coauthors, 1992: A comprehensive meteorological modeling system—RAMS. Meteor. Atmos. Phys., 49 , 6991.

  • Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Kluwer Academic, 666 pp.

  • Tremback, C. J., , J. Powell, , W. R. Cotton, , and R. A. Pielke, 1987: The forward-in-time upstream advection scheme: Extension to higher orders. Mon. Wea. Rev., 115 , 540555.

    • Search Google Scholar
    • Export Citation
  • Walko, R. L., and Coauthors, 2000: Coupled atmosphere–biophysics–hydrology models for environmental modeling. J. Appl. Meteor., 39 , 931944.

    • Search Google Scholar
    • Export Citation
  • Yoshino, M. M., 1975: Climate in a Small Area: An Introduction to Local Meteorology. University of Tokyo Press, 549 pp.

  • View in gallery

    Canopy components and heat fluxes. The free atmosphere, canopy air, vegetation, building, and ground (soil) are represented by the stylized letters A, C, V, B (where R denotes building roof and W denotes building walls), and G, respectively.

  • View in gallery

    Direct solar radiation SD incident on the horizontal surface in the UC model. Here, lground and lshadow represent the normalized ground width and the normalized shadow length on the ground, respectively; h denotes the building height; and θz indicates the solar zenith angle.

  • View in gallery

    Horizontal domains.

  • View in gallery

    (a) Land-use and (b) topography maps for grid 3.

  • View in gallery

    Diurnal cycle of the anthropogenic heat emission per unit area of urban land surface.

  • View in gallery

    Near-surface temperatures and wind simulated using the models and observed at the stations at (left) 0600 and (right) 1800 UTC 26 Jul 2001. Shown are the air temperatures at 2 m and surface wind simulated using (a) the original RAMS and (b) RAMS-UC and (c) the observed temperatures at 1.5 m and surface wind.

  • View in gallery

    Observed (squares) and simulated [triangles for the original RAMS without the urban canopy model (RAMS-org) and circles for the RAMS-UC model] near-surface air temperatures (solid lines) and relative humidities (dotted lines) at five surface observation sites located in grid 3.

  • View in gallery

    As in Fig. 7, but for the other two observation sites, located in grid 2 but outside grid 3.

  • View in gallery

    Locations of the observation stations.

  • View in gallery

    Scatterplot of the observed and simulated near-surface air temperatures.

  • View in gallery

    Near-surface air temperatures 2 m above ground (solid lines) and relative humidities (dashed lines) at a rural location (29.461°N, 106.444°E), simulated using the original RAMS model (triangles), the RAMS-UC model (circles), and the RAMS-UC model but with the anthropogenic heat not included (plus and times signs).

  • View in gallery

    As in Fig. 11, but for an urban location (29.569°N, 106.453°E).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 49 49 8
PDF Downloads 11 11 2

Modified RAMS-Urban Canopy Model for Heat Island Simulation in Chongqing, China

View More View Less
  • 1 Department of Urban Engineering, University of Tokyo, Tokyo, Japan
  • | 2 Institute of Observational Research for Global Change, Japan Agency for Marine–Earth Science and Technology, Yokosuka, Kanagawa, Japan
  • | 3 Department of Geography, Tokyo Metropolitan University, Tokyo, Japan
  • | 4 Department of Urban Engineering, University of Tokyo, Tokyo, Japan
© Get Permissions
Full access

Abstract

A single-layer urban canopy model was integrated into a nonhydrostatic meteorological model, the Regional Atmospheric Modeling System (RAMS). In the new model, called RAMS-Urban Canopy (RAMS-UC), anthropogenic heat emission was also considered. The model can be used to calculate radiation, heat, and water fluxes in an urban area, considering the geometric structure and thermodynamic characteristics of the urban canopy. The urban canopy was represented by normalized street canyons of infinite length, which were bordered by buildings on both sides. The urban region was covered by three types of surfaces: roof, wall, and road. Anthropogenic heat was emitted from these surfaces. Sensitivity tests between the original RAMS and the modified one were carried out by simulating the urban heat island (UHI) of Chongqing, located in an inland mountainous region in China. The results of the model were also compared with the observational data. It was found that the original model could not accurately simulate the UHI, in particular at night, whereas the accuracy was significantly improved in the RAMS-UC. The improvement is substantial even when anthropogenic heat emission is set to zero.

* Current affiliation: Department of Environmental Engineering, Building Research Institute, Tsukuba, Japan

Corresponding author address: Hongbin Zhang, Dept. of Environmental Engineering, Building Research Institute, 1 Tachihara, Tsukuba-shi, Ibaraki-ken 305-0802, Japan. Email: zhang@kenken.go.jp

Abstract

A single-layer urban canopy model was integrated into a nonhydrostatic meteorological model, the Regional Atmospheric Modeling System (RAMS). In the new model, called RAMS-Urban Canopy (RAMS-UC), anthropogenic heat emission was also considered. The model can be used to calculate radiation, heat, and water fluxes in an urban area, considering the geometric structure and thermodynamic characteristics of the urban canopy. The urban canopy was represented by normalized street canyons of infinite length, which were bordered by buildings on both sides. The urban region was covered by three types of surfaces: roof, wall, and road. Anthropogenic heat was emitted from these surfaces. Sensitivity tests between the original RAMS and the modified one were carried out by simulating the urban heat island (UHI) of Chongqing, located in an inland mountainous region in China. The results of the model were also compared with the observational data. It was found that the original model could not accurately simulate the UHI, in particular at night, whereas the accuracy was significantly improved in the RAMS-UC. The improvement is substantial even when anthropogenic heat emission is set to zero.

* Current affiliation: Department of Environmental Engineering, Building Research Institute, Tsukuba, Japan

Corresponding author address: Hongbin Zhang, Dept. of Environmental Engineering, Building Research Institute, 1 Tachihara, Tsukuba-shi, Ibaraki-ken 305-0802, Japan. Email: zhang@kenken.go.jp

1. Introduction

In comparison with global warming, the urban-warming phenomenon is much more evident and rapid. For example, the average annual temperature has increased by approximately 3°C during the past 100 yr for Tokyo and by about 2°–3°C for other major cities in Japan (Bureau of Environment 2007; Ministry of the Environment 2004), whereas the mean global temperature has increased by only 0.3°–0.6°C during the same period (Environmental Protection Agency 2005). Urban warming, the so-called urban heat island (UHI) effect, was first observed in London, England, by Luke Howard in the early eighteenth century and has since been reported in crowded and highly urbanized cities all over the world (e.g., Chandler 1965; Yoshino 1975; Landsberg 1981). Urban warming not only damages the amenities of cities, but also increases energy demand for air conditioning, degrades air quality, adversely affects urban climate, and even damages human health. One goal of contemporary urban environmental master planning is to control the UHI. Although various mitigation options, such as increasing vegetation, have been proposed, the effectiveness of countermeasures is yet to be clarified, a fact that is partly due to the lack of reliable simulations for comprehensive evaluation.

In meteorology, the horizontal scale from 2 to 2000 km is called the mesoscale, and that of less than 2 km is called the microscale (Orlanski 1975). Mesoscale models simulate the regional urban climate of a city in question, taking into account meteorological interactions over the entire city and the surrounding regions. Thus, they are often used to estimate the effect of the UHI—in particular, in discussing mitigation strategies for the master plan. However, most of the mesoscale models that are used for such purposes are based on hydrostatic equations, causing various problems, such as the limitation of the gridcell size, a low ability to simulate the boundary layer, and a degradation of accuracy in the vertical direction (e.g., Kondo 2001). In the past few years, nonhydrostatic models have been attracting the attention of urban climate researchers. In addition, the urban canopy and anthropogenic heat emission have become key issues in mesoscale UHI studies. The canopy layers usually refer to the layers between the ground surface and the tops of trees or buildings. They are called the vegetation canopy and the urban canopy, respectively. Conventional mesoscale atmospheric models simulate the urban canopy with the same scheme as that for the vegetation canopy, except that the soil constants and some other parameters are varied. The effects of the geometric structure of the urban canopy and the heat emission caused by human activities on the local meteorological behavior are ignored in such models. It has been pointed out that the impact of the urban canopy on the atmosphere cannot be accurately estimated using this method. In the past few years, numerous urban canopy models have been developed to represent the relationship between the street canyon geometry and the atmosphere (Kondo et al. 1999; Masson 2000; Kusaka et al. 2001; Martilli et al. 2002). In terms of their vertical structure, these urban canopy models can generally be divided into two groups: a single-layer approach and a multilayer approach. The single-layer model (e.g., Masson 2000; Kusaka et al. 2001) neglects the diversity of building heights and does not take into consideration the vertical inhomogeneity of physical characteristics within the urban canopy. On the other hand, the multilayer model (e.g., Kondo et al. 1999; Martilli et al. 2002) generally requires a large computational capacity to simulate the vertical structure of the urban canopy. Thus, it is not realistic to simulate the UHI using the multilayer model when the spatial scale of the city is large. In particular, it is difficult to estimate the UHI effect considering its relationship to the large-scale circulations in the surrounding regions. The design of the urban canopy (UC) model introduced in this paper is based on those of Kusaka et al. (2001) and Kusaka and Kimura (2004a, b). The Kusaka scheme is similar to that of Masson (2000). However, it considers the canyon orientation and the diurnal change in the solar azimuth angle and consists of several canyons with different orientations, as described by Kusaka et al. (2001). In addition, the authors’ UC model takes into consideration the effect of the anthropogenic sensible and latent heat fluxes on all artificial surfaces (building roofs, walls, and ground surfaces). This model can greatly contribute to the evaluation of various countermeasures to UHI. In addition, the UC model includes a maximum of 10 urban classes with different urban canyon structures to take into account the climatological differences among different districts in a city (e.g., commercial district and residential district).

Alternating the materials and shapes of buildings is one of the most effective methods of mitigating the UHI. Moreover, a reduction in anthropogenic heat is also important, because it is actually comparable to solar radiation in some regions and plays a significant role in the nighttime UHI. Thus, a highly accurate comprehensive modeling system considering the geometry of the urban canopy and anthropogenic heat emission is required to understand the processes of and to propose effective mitigation strategies for the UHI. The objective of this paper is to report the simulated results obtained using a relatively high resolution nonhydrostatic meteorological model, the Regional Atmospheric Modeling System (RAMS), coupled with a single-layer UC model. The modified model is applied to clarify the current UHI condition of an inland megacity, Chongqing, China, which is the key city in the national strategy to develop western China. The model is expected to contribute to future forecasting of UHI conditions in the city and also to enable the assessment of the impact of various mitigation methods.

2. Models

a. Meteorological model

The meteorological model used in this coupling experiment is version 4.4 of RAMS. RAMS is a highly versatile numerical code invented in the early 1980s at Colorado State University and is most frequently applied to the simulation of meteorological phenomena on the mesoscale (Tremback et al. 1987; Pielke et al. 1992; Cotton et al. 2003). Because the atmospheric model of RAMS is based on a full set of nonhydrostatic, compressible equations that take into account atmospheric dynamics and thermodynamics, as well as conservation equations for scalar quantities such as water vapor, liquid water, and ice mixing ratios, there is no lower limit to the gridcell size, and its two-way nesting function allows grid systems with different spatial resolutions to interact with each other. In this research, coarser grids are used to model the broad atmospheric environment and to provide boundary conditions for the nested grids inside them, and the highest-resolution grid is primarily used to analyze the UHI in the surface layer. The turbulence closure and radiation options configured in this study are the Mellor–Yamada level-2.5 scheme and the Chen and Cotton longwave–shortwave scheme, respectively (Cotton et al. 2003). RAMS also provides a vegetation canopy submodel called the Land Ecosystem–Atmospheric Feedback Model, version 2, (LEAF-2; Walko et al. 2000) to evaluate the heat and moisture exchanges between the atmosphere and the surfaces (canopy air, vegetation, soil, and snow). However, in urban land use, the LEAF-2 submodel can only compute the exchange by varying the soil constants and surface parameters (e.g., heat capacity, thermal conductivity, albedo, and roughness). It does not take into consideration the geometric structure of the urban canopy or the fluid dynamic and thermodynamic properties associated with a peculiar geometry, although this is one of the main causes of the UHI.

b. Urban canopy model

The design of the UC model coupled with RAMS in this research is based on the single-layer urban canopy model developed by Kusaka et al. (2001) and modified by Kusaka and Kimura (2004a, b). The original single-layer urban canopy model has the following five characteristics: 1) two-dimensional, symmetrical street canyons of infinite length to represent the geometric structure of the urban canopy, 2) the reflection of radiation and the role of building shadows, 3) the street canyon orientation and diurnal variation of the solar azimuth angle, 4) the orientation of the canyons to accommodate the change in the direction of solar radiation, and 5) a multilayer heat conduction scheme for building walls, roofs, and ground surfaces. The sensitivity of this single-layer urban canopy model (Kusaka scheme) was reported by Kusaka et al. (2001) as being similar to that of the complex multilayer models for heat and water fluxes, whereas its number of tunable variables is small and its parameterization is relatively simple. Hence, the computing methods of the Kusaka scheme were introduced into RAMS to represent the effect of the urban canopy on the thermal environment.

In addition, the Kusaka scheme does not consider the impact of anthropogenic heat, which is a significant contributor to the UHI. Therefore, anthropogenic heat was considered as comprising sensible and latent heat fluxes injected directly into the canopy air or free atmosphere from all artificial surfaces (ground areas and building walls and roofs), and its intensity varied with time.

c. Coupling method

In the RAMS model, the surface grid cells are divided into multiple subgrid patches, each with a different land use (five patches in this study), and the heat and moisture fluxes are computed separately at each patch using the LEAF-2 submodel. Therefore, a patch–land use diagnosis program is inserted when the RAMS initialization is started. If the land use of a certain patch is diagnosed as urban, the LEAF-2 submodel is replaced with the UC submodel in that patch. The UC submodel regards an urban class as an entirely artificial space consisting of numerous parallel street canyons (roads bordered by facing buildings). Each street canyon has the same geometric structure, in which its length is assumed to be much greater than its width. In each urban class, building height, building and road widths, and other surface parameters are set as constant, but they differ among urban classes. In the study presented here, owing to the lack of detailed land use data of the city, only two urban classes (urban and towns) are set in the UC submodel. In general, the maximum number of urban classes is 10 in the UC submodel. The UC submodel uses the bulk method to estimate the heat, moisture, and momentum fluxes inside the urban canopy. The submodel calculates the fluxes between the urban canopy and the overlying free atmosphere based on Monin–Obukhov similarity theory. The similarity function suggested by Dyer and Hicks (1970) is applied for unstable conditions, and Kondo et al.’s (1978) scheme is applied for stable cases. When urban patches coexist with vegetation patches in the same grid cell, the net heat and moisture fluxes from the surface to the free atmosphere are estimated as the sum of the results of the LEAF-2 and UC submodels calculated for each patch inside the grid cell.

The components and fluxes defined in a LEAF-2 patch and a UC patch are shown in Fig. 1. In the LEAF-2 patch, the canopy layer (C) beneath an atmospheric column (A) has partial vegetation cover (V). On the other hand, the UC patch includes building roofs (R), building walls (W), and ground surfaces (G). Note that G in the LEAF-2 patch denotes the soil layer. The arrows in Fig. 1 represent the fluxes inside canopy layers and those from the atmosphere to the canopy layers; the sans-serif H, W, and R in the figure indicate heat, water, and radiation fluxes, respectively. In Fig. 1, the first and second subscripts denote the source and receptor of the flux, respectively: s represents soil, v is vegetation, a is the free atmosphere, c is canopy air, r is building roofs, w is building walls, and g is ground surfaces in the urban canopy. However, Wsc′ does not indicate the water flux from the canopy to the soil associated with direct evaporation on the soil surface, but instead represents the water flux from the soil to the canopy air via vegetation. Numbered components in Fig. 1 (e.g., R1, R2) indicate that the model contains multiple layers for that component, but, for simplicity, only two layers are shown in the figure. The variables h and lroof denote the building height and width in the UC model, respectively. Note that the longwave radiation inside the urban canopy is not shown in Fig. 1. In the UC submodel, the heat and water are transported from the building walls and the ground surfaces to the canopy air, whereas those from building roofs go directly to the free atmosphere. The basic energy balance equation for individual fluxes from building walls, roofs, and ground surfaces is
i1558-8432-47-2-509-e1
where Rn,i is the net radiation flux to the surface, Hi is the sensible heat flux, (LE)i is the latent heat flux, and Gi is the ground heat flux.

d. Radiative fluxes

The calculation method of radiative fluxes is one of the largest differences between the LEAF-2 submodel and the UC submodel. The radiation scheme inside the urban canopy is very complex because of the existence of the buildings. The buildings reflect incident radiation, intercept direct solar radiation to the surface by forming shadows, absorb heat when the surrounding temperature is higher than that of their surfaces, and release it when the surrounding temperature is lower. Therefore, it is unsuitable to represent radiation fluxes around buildings by constant heat characteristics.

According to the Kusaka scheme, urban surfaces can be categorized into three types: building walls, building roofs, and ground areas. Their solar and longwave radiation fluxes are resolved individually, with the normalized shadow lshadow on the ground defined as
i1558-8432-47-2-509-e2
where h and lground are the normalized building height and ground width between building walls, respectively, and lroof is the normalized roof width (Fig. 2). In addition, θz is the solar zenith angle and θn is the angle between the horizontal direction of the solar radiation and the axis of the street canyon, which can be calculated from the solar azimuth angle and the canyon orientation. In addition, the orientations of street canyons in actual cities are not homogeneous in a grid cell. Therefore, the radiation budget should be calculated for various orientations and summed. In this UC model, the radiation budget is calculated eight times, rotating the street canyon by π/8 in turn. Then, the mean of the eight calculations is used.

The longwave radiation of building roofs, walls, and ground areas is considered in a similar way on the basis of the equations for the downward atmospheric longwave radiation, the surface temperature, and the surface view factor. The equations used in this part of the UC submodel are the same as those proposed by Kusaka et al. (2001).

e. Heat fluxes

Heat fluxes from building roofs, walls, and ground areas are calculated at each surface individually. Because the flux from building roofs is released to the free atmosphere directly, Monin–Obukhov similarity theory is used. However, the Jurges formula (Kusaka et al. 2001) is applied to building walls and ground surfaces because Monin–Obukhov similarity theory is not valid in the internal region of the urban canopy.

To simplify the anthropogenic heat calculation, it is assumed that all of the anthropogenic heat is emitted from the building walls, roofs, and ground surfaces in proportion to their surface areas. At this time, the diurnal cycle of anthropogenic heat can be set at hourly intervals by the user. Thus, sensible heat fluxes from the surfaces of building walls (HW) and ground areas (HG) are given as
i1558-8432-47-2-509-e3
i1558-8432-47-2-509-e4
i1558-8432-47-2-509-e5
where TS is the replaced canyon surface temperature defined at zT + d (zT is the roughness length for heat and d is the height of the zero-plane displacement); CW and TW are the bulk transfer coefficient and the surface temperature of the building walls, respectively; and CG and TG are those of the ground surfaces, respectively. The variable US is the wind speed in the urban canopy, and AHW and AHG are the anthropogenic heat emissions from the building walls and ground surfaces, respectively.
Because the heat capacity of the canopy air is negligible, the sensible heat flux released from the canopy air to the atmospheric layer (Ha) must be equal to the heat received by the canopy air from the ground and building walls:
i1558-8432-47-2-509-e6
If both urban and vegetated areas exist in one grid cell, the net sensible heat flux of that grid cell (H) is given as
i1558-8432-47-2-509-e7
where Au and Aυ are the ratios of the areas of the urban and vegetation surfaces to the total area, respectively; HR is the sensible heat flux from building roofs; AHR is the anthropogenic heat from building roofs, and Hυ is the sensible heat flux from the vegetation surface, which is calculated by the LEAF-2 submodel.
The latent heat fluxes for building walls [(LE)W] and ground areas [(LE)G] are computed as
i1558-8432-47-2-509-e8
and
i1558-8432-47-2-509-e9
where β and C represent the availability of moisture and the transfer coefficient at the artificial surface in the urban canopy, respectively, and cp denotes the specific heat capacity of dry canopy air at a constant pressure. The variable qsat(T) is the saturated specific humidity at the temperature T of the surface, qc is the specific humidity of the canopy air, and ALE indicates the anthropogenic latent heat. Subscripts G and W attached to β, C, T, and ALE represent the building walls and the ground surfaces, respectively.

3. Simulation

In this study, the current heat island condition in Chongqing was simulated using the modified RAMS-UC model, and a reference run was performed using the original RAMS model without any modification. A simulation using the RAMS-UC model without anthropogenic heat emission was also performed to investigate the contribution of the urban canopy itself and that of the anthropogenic heat. The performances of the two models were evaluated by comparing their results with data from local observations provided by the Chongqing Weather Bureau. The observation data included temperature, wind direction (16 quarters) and speed, and relative humidity at seven ground observation stations. The observed data from 0000 UTC 26 July to 0000 UTC 27 July 2001 were analyzed for comparison. This period was selected as a warm sunny day.

a. Study area

Chongqing municipality is situated in the southeastern part of the Sichuan basin and on the upper reaches of the Yangtze River at the head of a reservoir behind the Three Gorges Dam. The main city area in Chongqing municipality is located where the Yangtze and Jialing Rivers meet and is nestled between two mountain chains. Its terrain is characterized by rugged hills, and it forms a typical inland mountain city. Winter in Chongqing is warm, and summer is extremely hot. Chongqing is one of the three famous “stoves” in China. The average maximum temperature in July is 34.4°C, and the number of days with a maximum temperature higher than 35°C is approximately 40 per year (China eFair 2005, unpublished manuscript). In 1997, to promote the economic development of western China, as well as to coordinate the relocation of the immigrants from the Three Gorges Dam project area, the original Chongqing City was promoted to the status of a municipality. Thus, the current Chongqing municipality is spread over a total area of 82 400 km2 and has a population of more than 30 million. Chongqing municipality is the beachhead for the development of the western part of China. Massive construction and development works are currently under way in the city. For example, the constructed urban area of the city of Chongqing is expected to increase from 415.5 km2 (in 2003) to 820 km2 by 2020, and the urban population is expected to expand from 5.2 million to 8.8 million in the same period. There is a strong chance that the UHI will become a serious problem in the near future in the city of Chongqing.

b. Domain information

Because the city of Chongqing has a large area and a complex topography, a wide range of meteorological phenomena should be taken into consideration for a simulation. The total domain area (grid 1) was 2000 km × 2000 km, with a center point at (29.58°N, 106.5°E), which is near the city center of Chongqing, and was divided into 60 × 60 grid cells. Two nested grids were set inside grid 1. The coarser nested grid (grid 2) had 60 × 60 cells with a grid cell size of 4 km × 4 km, and the finer nested grid (grid 3) had 80 × 80 cells with a grid cell size of 1 km × 1 km (Fig. 3).

The vertical domain consisted of two parts: soil and atmosphere. The underground part had 11 effectual layers, whose depths were given as 1, 3, 6, 9, 12, 16, 20, 25, 30, 40, and 50 cm. In the terrain-following σz coordinate system used in RAMS, the vertical domain assumes a perfectly flat top and a terrain-following bottom, and its coordinates are defined as
i1558-8432-47-2-509-e10
where H is the height of the vertical domain top and zg is the local topographic height. Because the horizontal domain includes part of the Tibetan Plateau, where the topographic height is more than 6000 m, the vertical domain must be sufficiently large to accommodate this. Therefore, the top of the vertical domain was set to approximately 20 000 m above sea level. The number of layers was 29 for grid 1, and 41 for grids 2 and 3. Here, a variable vertical grid-spacing method supported by RAMS was applied to increase the resolution near the ground. The thickness of the bottom layer in grid 1 was 120 m; for layers above the bottom layer, the layer thickness increased by 15 from the layer below. When the calculated thickness of a layer exceeded 1000 m, the thickness was set to 1000 m. In the cases of grids 2 and 3, because cities are expected to affect the lowest several hundred meters of the atmosphere (e.g., Ichinose et al. 1999), each of the lowest six layers (up to approximately 1000 m) was divided further into four, four, three, three, two, and two sublayers, respectively, to resolve more details near the ground.

c. Input data

RAMS surface characteristic datasets (including topography and land use) and National Centers for Environmental Prediction (NCEP) data were used as the surface data for grids 1 and 2 and for the meteorological input, respectively. For the topography data and land-use input data for grid 3, revised Advanced Spaceborne Thermal Emission and Reflection Radiometer Digital Elevation Model (ASTER DEM) data and the data extracted from Landsat-7 satellite images were applied and interpolated with grid 2. A summary of the applications and resolutions of each dataset is shown in Table 1, and the topography and land-use maps for Grid 3 are shown in Fig. 4.

d. Parameter setting

In the UC submodel, the building heights for urban areas and towns were set as 12 and 4 m, respectively, which were appropriate for the average building heights in the city of Chongqing and its surrounding towns. The artificial surfaces (building walls, roofs, and ground area) were also composed of multiple layers. Each surface consisted of 11 layers. The total thickness of the ground area was 330 cm, whereas that of building walls and roofs was 146.3 cm. The heat fluxes among the internal layers were calculated using a one-dimensional energy conservation equation.

Owing to the lack of statistics, there was no reliable heat discharge data for buildings or roads in the city of Chongqing. Hence, in this study, the sensible anthropogenic heat was estimated on the basis of total energy consumption in the city of Chongqing, whereas the latent anthropogenic heat was ignored. According to the Chongqing Statistical Yearbook 2006 (Chongqing Municipal Bureau of Statistics 2007), the energy consumption of the entire city was equivalent to approximately 25 × 109 kg of coal in 2001, and approximately 70% of this was consumed in the urban area. Dividing 70% of the total annual energy consumption by 365 days, the daily energy consumption per unit area of the surface was assumed to be approximately 4 million J m−2 day−1 (calorific power of 1 kg coal equivalent = 29.3 MJ). Then, all consumed energy was assumed to be released as sensible heat to the surrounding air. In consideration of hourly fluctuations in energy demand, the magnitude of sensible heat discharge per unit area was set at 70 W m−2 from 0800 to 1900 local Beijing time, at 40 W m−2 from 0600 to 0800 LT and from 1900 to 2200 LT, at 20 W m−2 from 0500 to 0600 LT and from 2200 to 0000 LT, and at 10 W m−2 from 0000 to 0500 LT. In the case of towns, the heat discharge magnitude per unit area was calculated as 20% of that in the urban area. It is assumed that all artificial surfaces in the same urban class release anthropogenic heat with the same area density (AHR = AHW = AHG). Under this assumption, the total amounts of anthropogenic heat released from building roofs, walls, and ground areas are only determined by the ratio of their normalized size parameters (Table 2). The diurnal cycle of the net anthropogenic sensible heat emission per unit area of the urban land surface AH = AHRlroof + AHGlroad + 2AHWh is shown in Fig. 5. The other main surface parameters for the UC and LEAF-2 submodels are listed in Tables 2 and 3, respectively.

4. Results and discussion

The simulations were performed using a single six-node Linux cluster (kernel 2.4, Red Hat Linux 9.0, PGI 6.0), for a simulation time period of 30 h with a time step of 12 s. The CPU time required for the original RAMS model was 36 h, and that for the RAMS-UC model was 37 h. The standard results, including the potential temperature, specific humidity, wind speed, heat fluxes, and so on, of every grid cell at each vertical layer, were recorded hourly. As an example, Fig. 6 illustrates the differences in the temperature distributions at 2 m above the ground, calculated using the RAMS-UC and original RAMS models. Figures 6a.1 and 6b show the near-surface air temperature distributions 2 m above the ground calculated using the original RAMS (RAMS-org) and RAMS-UC models, respectively, at 0600 UTC (1400 Beijing time—Beijing time is used as the local standard time at Chongqing; however, the local time at Chongqing is delayed by approximately 1 h from Beijing time), whereas Figs. 6a.2 and 6b show those at 1800 UTC (0200 Beijing time). To compare the simulation results and observations, the observed wind direction and speed were indicated in these four distribution maps as large arrows. Also, the observed near-surface air temperatures at five stations inside grid 3 at 0600 and 1800 UTC are shown in Figs. 6c.1 and 6c, respectively.

In these two simulations, all of the parameters were the same. As shown in Fig. 6, the wind direction in both models showed trends similar to those of the observational data, and there was no significant difference between the two models, although the wind speed around the urban area was slightly higher in the RAMS-UC model than in the original RAMS model. Moreover, the RAMS-UC model was successful in expressing the effects of the urban area on the air temperature near the ground as compared with the original RAMS model. In the RAMS-UC model, the UHI is more intense at nighttime than during the daytime, as shown in Figs. 6b.1 and 6b. The air temperature 2 m above the ground of the urban area calculated using the RAMS-UC model was nearly 0.5°C higher than that of the original RAMS model in the daytime, whereas it was nearly 6°C higher at nighttime at maximum. These differences were slightly larger than those for plain or coastal cities (e.g., about 3°C in Tokyo; Ministry of the Environment 2004). In the case of the city of Chongqing, the downslope drainage of air at nighttime is prevented by the rugged terrain, resulting in little impact in cooling the city. Thus, the mountain ranges on both sides of the city of Chongqing act as a barrier of heat dispersion much more than they alleviate the UHI. This phenomenon suggests a possible problem in the mitigation of the UHI in cities located in mountainous regions such as Chongqing.

To examine the difference in the depth of the daytime planetary boundary layer (PBL) in both models, the vertical profiles of potential temperature at 0600 UTC were examined at the center of grid 3, where the center of the UHI is located. The depth of the mixed layer is approximately 800 m in the original model, whereas it reaches 1200 m in the RAMS-UC. In addition, the neutral stratification in the mixed layer is more clearly identified in the RAMS-UC (not shown). The mixed layer becomes thick because of stronger heating near the surface. It is inferred that the vertical mixing is enhanced by the thicker mixed layer. The wind speed in the lowest layer of the atmosphere is also higher in the RAMS-UC, as indicated in Fig. 6. The vertical shear of the wind in the mixed layer is small (not shown). It is well known that the near-surface wind becomes stronger in an unstable case relative to a neutral or stable case because of stronger vertical mixing (e.g., Stull 1988). It is suggested that the higher near-surface wind speed over the urban region in RAMS-UC is associated with larger instability caused by a strengthened heat island.

To evaluate the performances of the original RAMS and RAMS-UC in a more detail, the two simulation results were compared with observations in the city of Chongqing (Figs. 7 and 8). The surface observation data used here are from the Chongqing Weather Bureau. These observations contain 6-hourly 1.5-m air temperature, relative humidity, and 10-m wind speed and direction (16 quarters) at seven ground observation stations. Among these stations, five stations were inside grid 3 and the other two stations were outside it (in grid 2). Their positions and classifications are shown in Fig. 9. Because the variation in air temperature near the ground accurately represented the UHI, the near-surface air temperature 2 m above the ground was mainly used to discuss the precision of the models in this study.

The thermometer shelters used at all seven ground observation stations were located 1.5 m above a grassy area of over 1000 m2. Because the smallest grid cell size was 1000 m × 1000 m in the simulations, the observed data in these green areas may not correspond well to the simulated result at each grid. Other papers also mentioned this problem when comparing simulated results with observations (e.g., Hadano et al. 2004). The simulated temperatures may be slightly higher than the data recorded from a thermometer shelter. However, it is difficult to assess this mismatch accurately. Nonetheless, the comparison of simulation results with surface observations can also verify the accuracy of the models.

The standard analytical files were obtained hourly, and air temperature results 2 m above the ground were estimated from the potential temperature and canopy temperature, whereas the relative humidity inside the canopy layer was assumed to be equal to the relative humidity in the bottom atmospheric layer. In Figs. 7 and 8, the solid lines denote the temperature, and dotted lines indicate the relative humidity. Lines with squares represent the observations, lines with triangles represent the simulated results of the original RAMS model, and lines with circles show the simulated results of the RAMS-UC model. Note that the results of the simulation represent the mean values of the grid cells inside which the observation points were located. The results shown in Fig. 7 were extracted from grid 3, and those in Fig. 8 were extracted from grid 2. Because the gridcell size of grid 3 was smaller (1:16) than that of grid 2, the accuracy of the results in Fig. 8 was much lower than that for those in Fig. 7. Hence, mainly stations inside grid 3 were used to evaluate the sensitivity of the models. A scatter diagram of the simulated results of the two models at these stations is shown in Fig. 10.

In Fig. 7, the maximum near-surface air temperature discrepancy between the results obtained by RAMS-UC and the surface observations (simulated results − observations, here and below) was +4.39°C, which appeared at the third station Yubei at 0600 UTC (1400 Beijing time). The maximum relative humidity discrepancy between them was −11%, which appeared at the fifth station Banan at 0000 UTC (0800 Beijing time). The root-mean-square errors of temperature and relative humidity were 0.36°C and 1.38%, respectively. For the original RAMS model, on the other hand, the maximum discrepancies of temperature and relative humidity were −6.24°C at 1200 UTC (2000 Beijing time) at the Banan station and +33% at 1800 UTC (0200 Beijing time) at the Jiangjin station, and the root-mean-square errors were 0.78°C and 4.80%, respectively. Although the discrepancy shown in Fig. 8 was larger than that in Fig. 7 for the reasons explained above, the results in Fig. 8 were similar to those in Fig. 7. The maximum difference and root-mean-square errors of the temperature and the relative humidity for the RAMS-UC model were +2.43°C, 0.55°C, −22%, and 2.55%, respectively, whereas those of the original RAMS model and the surface observations were −5.01°C, 1.19°C, +41%, and 8.65%, respectively.

It was demonstrated that both models could be used to describe the daytime near-surface air temperature relatively well, by referring to the actual near-surface air temperature at each observational time. However, it was found that the original RAMS model underestimated the nighttime temperature by about 4.0°C, whereas the RAMS-UC model reduced this difference to less than 0.5°C, on average, at five stations. The current study suggests that the original RAMS model underestimates the magnitude of the UHI—in particular, at night—and that the RAMS-UC model overcomes this problem. On the other hand, the relative humidity decreases corresponding to the temperature increase at nighttime, although the specific humidity is not much different between the two models. In the daytime, however, the difference in the temperature is not large, and the relative humidity decreases because of lower moisture availability in the RAMS-UC model.

Figures 11 and 12 show the hourly near-surface air temperatures 2 m above the ground and the relative humidities in the bottom layer at a rural location and an urban location, respectively, from 30-h simulations using RAMS-UC (circles), RAMS-UC without anthropogenic heat emission (plus and times signs), and the original RAMS (triangles). By comparing the results simulated using RAMS-UC at the rural (Fig. 11) and urban (Fig. 12) locations, it was confirmed that the UHI intensity is high in the afternoon and at night. The results also showed that there was no clear difference between the results of the original RAMS and RAMS-UC models at the rural location. However, at the urban location, the RAMS-UC model showed a much higher temperature than that shown by the original RAMS. The largest discrepancy appeared at approximately 1600 Beijing time, 14 h after the simulation started, when the typical urban climate is most clearly observed. If one ignores the results from the first 6 h after the simulation was started, the nighttime (0000–0600 Beijing time, simulation time 23–29 h) temperatures calculated using the RAMS-UC model were, on average, 4°C higher than those calculated using the original RAMS model. This contributes to the improvement of the simulation of the UHI during the nighttime. By comparing the results from the RAMS-UC and those without anthropogenic heat emission, it was found that the contribution of the urban canopy itself is larger than that of anthropogenic heat. It is suggested that we should focus on the existence of buildings itself rather than on anthropogenic heat emission when we consider the mitigation of the urban heat island.

5. Conclusions

A modified version of RAMS was developed by incorporating an urban canopy submodel into the original RAMS. The simulated results for the UHI in Chongqing were verified by comparing them with the observed data. Results showed that the modified RAMS-UC model simulated the UHI more accurately than did the original RAMS model. In the finest grid system, the maximum discrepancy of air temperature at the 2-m level was reduced from 6.4° to 4.4°C. Moreover, the root-mean-square error decreased from 0.8° to 0.4°C.

In the original RAMS, the air temperature near the surface mostly depended on the topography. However, the model underestimated the effects of land use and the geometric structure of the urban area, and it ignored anthropogenic heat emission. It was found that the urban canopy submodel reduced these shortcomings while the merits of the original RAMS model, such as the nonhydrostatic equations and the two-way nesting system, were maintained. The RAMS-UC model was shown to describe the meteorological conditions near the surface with a high resolution, considering realistic boundary conditions.

Chongqing is currently undergoing rapid expansion and reconstruction and is facing intensive UHI effects. From a proactive standpoint, the improved UHI modeling approach is expected to be a useful tool for the city as it pursues its development in a more sustainable manner.

Acknowledgments

We acknowledge Atmospheric, Meteorological, and Environmental Technologies (ATMET) for providing the source code for the RAMS model. We are grateful for the support provided by Chongqing University. This paper was financially supported by a grant-in-aid for scientific research (A) by the Japan Society for the Promotion of Science (JSPS).

REFERENCES

  • Bureau of Environment, cited. 2007: The environment of Tokyo 2006. Tokyo Metropolitan Government Environmental White Paper 2006. [Available online at https://www2.kankyo.metro.tokyo.jp/kouhou/env/eng_2006/index.html.].

  • Chandler, T. J., 1965: : The Climate of London. Hutchinson, 292 pp.

  • Chongqing Municipal Bureau of Statistics, cited. 2007: Chongqing Statistical Yearbook 2006. [Available online at http://www.cqtj.gov.cn/tjnj/2006/yearbook/indexe.htm.].

  • Cotton, W. R., and Coauthors, 2003: RAMS 2001: Current status and future directions. Meteor. Atmos. Phys., 82 , 529.

  • Dyer, A. J., , and B. B. Hicks, 1970: Flux gradient relationship in the constant flux layer. Quart. J. Roy. Meteor. Soc., 96 , 715721.

  • Environmental Protection Agency, cited. 2005: Global warming—Climate. Web page of the U.S. Environmental Protection Agency. [Available online at http://yosemite.epa.gov/oar/globalwarming.nsf/content/climate.html.].

  • Hadano, K., , T. Izumi, , D. Nakayama, , and H. Matsuyama, 2004: The spatial representativity of temperature data of AMeDAS as revealed by the variogram. (in Japanese). Theor. Appl. GIS, 12 , 3546.

    • Search Google Scholar
    • Export Citation
  • Ichinose, T., , K. Shimodozono, , and K. Hanaki, 1999: Impact of anthropogenic heat on urban climate in Tokyo. Atmos. Environ., 33 , 38973909.

    • Search Google Scholar
    • Export Citation
  • Kondo, H., 2001: Simulation in the field of atmospheric environment. (in Japanese). J. Comput. Fluid Dyn., 9 , 133140.

  • Kondo, H., , Y. Kikegawa, , Y. Genchi, , and S. Yamamoto, 1999: Heating in the urban canopy by anthropogenic heat use. Proc. 15th Int. Congress of Biometeorology and International Conf. on Urban Climatology, Sydney, NSW, Australia, International Society of Biometeorology (ISB) and World Meteorological Organization (WMO), CD-ROM, 249.

  • Kondo, J., , O. Kanechika, , and N. Yasuda, 1978: Heat and momentum transfers under strong stability in the atmospheric surface layer. J. Atmos. Sci., 35 , 10121021.

    • Search Google Scholar
    • Export Citation
  • Kusaka, H., , and F. Kimura, 2004a: Coupling a single-layer urban canopy model with a simple atmospheric model: Impact on urban heat island simulation for an idealized case. J. Meteor. Soc. Japan, 82 , 6780.

    • Search Google Scholar
    • Export Citation
  • Kusaka, H., , and F. Kimura, 2004b: Thermal effects of urban canyon structure on the nocturnal heat island: Numerical experiment using a mesoscale model coupled with an urban canopy model. J. Appl. Meteor., 43 , 18991910.

    • Search Google Scholar
    • Export Citation
  • Kusaka, H., , H. Kondo, , Y. Kikegawa, , and F. Kimura, 2001: A simple single-layer urban canopy model for atmospheric models: Comparison with multi-layer and slab models. Bound.-Layer Meteor., 101 , 329358.

    • Search Google Scholar
    • Export Citation
  • Landsberg, H. E., 1981: The Urban Climate. Vol. 28. Academic Press, 275 pp.

  • Martilli, A., , A. Clappier, , and M. W. Rotach, 2002: An urban surface exchange parameterization for mesoscale models. Bound.-Layer Meteor., 104 , 261304.

    • Search Google Scholar
    • Export Citation
  • Masson, V., 2000: A physical-based scheme for the urban energy budget in atmospheric models. Bound.-Layer Meteor., 94 , 357397.

  • Ministry of the Environment, cited. 2004: Outline of the policy framework to reduce urban heat island effects. Inter-Ministry Coordination Committee to Mitigate Urban Heat Island, Ministry of the Environment, Japan. [Available online at http://www.env.go.jp/en/air/heat/heatisland.pdf.].

  • Orlanski, I., 1975: A rational subdivision of scales for atmospheric processes. Bull. Amer. Meteor. Soc., 56 , 527530.

  • Pielke, R. A., and Coauthors, 1992: A comprehensive meteorological modeling system—RAMS. Meteor. Atmos. Phys., 49 , 6991.

  • Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Kluwer Academic, 666 pp.

  • Tremback, C. J., , J. Powell, , W. R. Cotton, , and R. A. Pielke, 1987: The forward-in-time upstream advection scheme: Extension to higher orders. Mon. Wea. Rev., 115 , 540555.

    • Search Google Scholar
    • Export Citation
  • Walko, R. L., and Coauthors, 2000: Coupled atmosphere–biophysics–hydrology models for environmental modeling. J. Appl. Meteor., 39 , 931944.

    • Search Google Scholar
    • Export Citation
  • Yoshino, M. M., 1975: Climate in a Small Area: An Introduction to Local Meteorology. University of Tokyo Press, 549 pp.

Fig. 1.
Fig. 1.

Canopy components and heat fluxes. The free atmosphere, canopy air, vegetation, building, and ground (soil) are represented by the stylized letters A, C, V, B (where R denotes building roof and W denotes building walls), and G, respectively.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 2.
Fig. 2.

Direct solar radiation SD incident on the horizontal surface in the UC model. Here, lground and lshadow represent the normalized ground width and the normalized shadow length on the ground, respectively; h denotes the building height; and θz indicates the solar zenith angle.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 3.
Fig. 3.

Horizontal domains.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 4.
Fig. 4.

(a) Land-use and (b) topography maps for grid 3.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 5.
Fig. 5.

Diurnal cycle of the anthropogenic heat emission per unit area of urban land surface.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 6.
Fig. 6.

Near-surface temperatures and wind simulated using the models and observed at the stations at (left) 0600 and (right) 1800 UTC 26 Jul 2001. Shown are the air temperatures at 2 m and surface wind simulated using (a) the original RAMS and (b) RAMS-UC and (c) the observed temperatures at 1.5 m and surface wind.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 7.
Fig. 7.

Observed (squares) and simulated [triangles for the original RAMS without the urban canopy model (RAMS-org) and circles for the RAMS-UC model] near-surface air temperatures (solid lines) and relative humidities (dotted lines) at five surface observation sites located in grid 3.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 8.
Fig. 8.

As in Fig. 7, but for the other two observation sites, located in grid 2 but outside grid 3.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 9.
Fig. 9.

Locations of the observation stations.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 10.
Fig. 10.

Scatterplot of the observed and simulated near-surface air temperatures.

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 11.
Fig. 11.

Near-surface air temperatures 2 m above ground (solid lines) and relative humidities (dashed lines) at a rural location (29.461°N, 106.444°E), simulated using the original RAMS model (triangles), the RAMS-UC model (circles), and the RAMS-UC model but with the anthropogenic heat not included (plus and times signs).

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Fig. 12.
Fig. 12.

As in Fig. 11, but for an urban location (29.569°N, 106.453°E).

Citation: Journal of Applied Meteorology and Climatology 47, 2; 10.1175/2007JAMC1397.1

Table 1.

List of input datasets.

Table 1.
Table 2.

Main surface parameters for the UC submodel.

Table 2.
Table 3.

Main surface parameters for the LEAF-2 submodel.

Table 3.
Save