Assessment of interpolated ERA-40 reanalysis temperature and precipitation against observations of the Balkan Peninsula

Due to their ready availability and temporal and spatial consistency, reanalysis data are widely used within the climate community. Nevertheless, higher spatial resolutions are often required and statistical interpolation techniques are applied to increase the data resolution. This work aims to derive a set of high spatial resolution data through three-dimensional interpolation of daily temperature and precipitation. Thin plate spline interpolation has been chosen and used to interpolate ERA-40 temperature and precipitation from a coarse grid (110 km) into a finer one of 1-km spatial resolution. The study evaluates the method by comparing the simulated variables with available in situ meteorological measurements. The chosen stations are distributed over the study region and, most importantly, contain information from a range of altitudes. The results indicate that accounting for the topography in the interpolation process improves the comparisons, with the biggest improvements being evident in the most mountainous areas. The method is found to be better in estimating temperature than precipitation fields. Moreover, the method performs better for maximum temperature in high altitudes and for minimum temperature in low altitudes.     

Cyclone track prediction by a quasi-Lagrangian limited area model

Summary. ?Cyclone track predictions in the Indian seas (Bay of Bengal and Arabian Sea) with a quasi-Lagrangian model (QLM) have been attempted. QLM has a horizontal resolution of 40 km and 16 sigma levels in the vertical. It is integrated in a domain of about 4400 × 4400 km². A new initialization procedure to provide initial fields for running the model has been designed. The initialization procedure consists of updating the global model forecasts, used as first guess, provided by the National Center for Medium Range Weather Forecasting (NCMRWF), New Delhi. A new version of IMD’s operational optimum interpolation scheme has been created to suit the QLM grid structure. Lateral boundary conditions are computed from the extended forecasts of NCMRWF. The track forecasts in each case show a reasonable skill of the forecast model in predicting the direction of movement within acceptable limits of forecast errors, which are comparable to some of the best models operated by advanced NWP centers of the world. Even the recurving storms are well predicted. Evolution of the vertical motion fields are also studied which reveal some interesting features, which are described in detail in the text. The composited vertical motion fields are projected against observed rainfall distribution, which show a good spatial correspondence. Received August 9, 2001; revised March 12, 2002; accepted June 17, 2002 Published online: May 8, 2003     

Local regression models for spatial interpolation of urban heat island—an example from Wroc?aw, SW Poland

Geographically weighted regression algorithm (GWR) has been applied to derive the spatial structure of urban heat island (UHI) in the city of Wroc?aw, SW Poland. Seven UHI cases, measured during various meteorological conditions and characteristic of different seasons, were selected for analysis. GWR results were compared with global regression models (MLR), using various statistical procedures including corrected Akaike Information Criterion, determination coefficient, analysis of variance, and Moran’s I index. It was found that GWR is better suited for spatial modeling of UHI than MLR models, as it takes into account non-stationarity of the spatial process. However, Monte Carlo and F3 tests for spatial stationarity of the independent variables suggest that for several spatial predictors a mixed GWR–MLR approach is recommended. Both local and global models were extended by the interpolation of regression residuals and used for spatial interpolation of the UHI structure. The interpolation results were evaluated with the cross-validation approach. It was found that the incorporation of the spatially interpolated residuals leads to significant improvement of the interpolation results for both GWR and MLR approaches. Because GWR is better justified in terms of statistical specification, the combined GWR?+?interpolated regression residuals (GWR residual kriging; GWRK) approach is recommended for spatial modeling of UHI, instead of widely applied MLR models.     

A description of persistent climatic anomalies in a 1000-year climatic model simulation

 The Mark 2 version of the CSIRO coupled global climatic model has been used to generate a 1000-year simulation of natural (i.e. unforced) climatic variability representative of “present conditions”. The annual mean output from the simulation has been used to investigate the occurrence of decadal and longer trends over the globe for a number of climatic variables. Here trends are defined to be periods of years with a climatic anomaly of a given sign. The analysis reveals substantial differences between the trend characteristics of the various climatic variables. Trends longer than 12 years duration were unusual for rainfall. Such trends were fairly uniformly distributed over the globe and had an asymmetry in the rate of occurrence for wet or dry conditions. On the other hand, trends in surface wind stress, and especially the atmospheric screen temperature, were of longer duration but primarily confined to oceanic regions. The trends in the atmospheric screen temperature could be traced deep into the oceanic mixed layer, implying large changes in oceanic thermal inertia. This thermal inertia then constituted an important component of the `memory' of the climatic system. While the geographic region associated with a given trend could be identified over several adjacent grid boxes of the model, regional plots for individual years of the trend revealed a range of variations, suggesting that a consistent forcing mechanism may not be responsible for a trend at a given location. Typical return periods for 12-year rainfall trends were once in 1000 years, highlighting the rarity of such events. Using a looser definition of a trend revealed that drying trends up to 50 years duration were also possible, attributable solely to natural climatic variability. Significant (∼20% to 40%) rainfall reductions per year can be associated with a long-term drying trend, hence such events are of considerable climatic significance. It can take more than 100 years for the hydrologic losses associated with such a trend to be overcome. Overall, the simulation provides new and useful insights into climatic trends, and quantifies a number of poorly observed characteristics. The results highlight the extensive and pervasive influence of unforced natural climatic variability as an omnipresent generator of climatic trends. Received: 20 January 2000 / Accepted: 21 September 2000     

Evaluating spatial scales of climate variability in sub-Saharan Africa

Summary Spatial scales of variability in seasonal rainfall over Africa are investigated by means of statistical and numerical techniques. In the statistical analysis spatial structure is studied using gridded 0.5° resolution monthly data in the period 1948–1998. The de-seasonalized time series are subjected to successive principal component (PC) analysis, allowing the number of modes to vary from 10 to 24, producing cells of varying dimension. Then the original rainfall data within each cell are cross-correlated (internal), then averaged and compared with the adjacent cells (external) for each PC solution. By considering the ratio of internal to external correlation, the spatial scales of rainfall variability are evaluated and an optimum solution is found whose cell dimensions are approximately 10⁶ km². The aspect of scale is further studied for southern Africa by consideration of numerical model ensemble simulations over the period 1985–1999 forced with observed sea surface temperatures (SSTs). The hindcast products are compared with observed January to March (JFM) rainfall, based on a station-satellite merged analysis of precipitation (CMAP) data at 2.5° resolution. Validations for different sized areas indicate that cumulative standardized errors are greatest at the scale of a single grid cell (10⁴ km²) and decrease 20–30% by averaging over successively larger areas (10⁶ km²).     

Quantifying spatial patterns of bioclimatic zones and controls in Turkey

Summary The study was aimed at inferring spatial patterns of climatic zones as well as identifying significant discriminating bioclimatic controls for distribution of major ecosystems in Turkey, based on multivariate analyses. A total of 12 climate variables and 11 bioclimatic indices for the period of 1968–2004 at 272 meteorological stations, and four location data (latitudes, longitudes, altitudes, and distance to sea) were analyzed using discriminant analysis (DA), hierarchical and non-hierarchical cluster analyses (CA), principal components analysis (PCA), and multiple linear regression (MLR) models. The first three and four linear discriminant functions (LDFs) explained 88 and 95% of the variation in the dataset, respectively. The efficacy of the discriminant model was high (85.5%) based on the cross-validation method. The hierarchical and non-hierarchical CA pointed to seven clusters (climate types) that can be observed on the basis of broad climatic similarity of 97%. PCA elucidated 78% of variation in the dataset. MLR models that accounted for variations in the 12 climatic response variables as a function of the four location variables and aspect had R ² values ranging from 28.8% for precipitation to 89.8% for mean air temperature and soil temperature for a depth of 5 cm. The multivariate analyses indicated that the meteorological stations are heterogeneous clusters consisting of the seven climatic zones. However, differences in the bioclimatic variables at the boundaries complicate the natural clustering scheme of a multidimensional cloud of data points and were detected in a climatologically plausible manner by the Ward and K-means CA, and PCA. Our multivariate approach revealed that the commonly used climatic zones are insufficient representations of the inferred climatic zones: (1) the coastal Black Sea; (2) the inland Black Sea; (3) the southeastern Anatolia; (4) the eastern Anatolia; (5) the central Anatolia; (6) the Mediterranean; and (7) the Aegean. Authors’ addresses: F. Evrendilek, Department of Environmental Engineering, Faculty of Engineering and Architecture, Abant Izzet Baysal University, G?lk?y Cambus, 14280 Bolu, Turkey; S. Berberoglu, Department of Landscape Architecture, Cukurova University, Balcali-Adana, Turkey.     

A global ocean temperature and altimeter data assimilation system for studies of climate variability

 An ocean data assimilation (ODA) system which can assimilate both temperature and altimeter observations has been applied to the global ocean and tested between January 1993–October 1996. A statistical method has been used to convert sea surface height (SSH) anomalies observations from TOPEX/POSEIDON into synthetic temperature profiles. The innovative aspect of this method is the introduction of time dependency in the correlations used to transform the altimeter observations into temperature corrections. The assimilation system is based on a univariate variational optimal interpolation scheme applied to assimilate both in situ and synthetic temperature profiles. In addition, a longer global analysis for the upper-ocean temperature starting from January 1979 and ending November 1997, has been produced to examine the skill of sea temperature assimilation with a rather simple and practical method. The temperature analysis shows encouraging improvement over a corresponding ocean simulation when compared to independent (not assimilated) temperature data both at seasonal and interannual time scales. However, the univariate data assimilation of hydrographic data does not result in an improvement of the velocity field. In fact the assimilation of sparse in situ data can introduce unrealistic spatial variability in the temperature field which affects the velocity field in a negative way. This deficiency is partially overcome when we also assimilate altimeter observations since the coverage is complete and uniform for this data. In particular, our study shows that temperature corrections due to the altimeter signal have a positive impact on the current system in the tropical Pacific. Received: 28 May 2000 / Accepted: 6 November 2000     

Solar Radiation Climatology of Alaska

Summary There are only six locations in Alaska for which global radiation data of more than a year in duration are available. This is an extremely sparse coverage for a state which covers 1.5×10&⁶ km² and stretches over at least three climatic zones. Cloud observations are, however, available from 18 stations. We used fractional cloud cover and cloud type data to model the global radiation and thus obtain a more complete radiation coverage for Alaska. This extended data set allowed an analysis of geographic and seasonal trends. A simple 1-layer model based on Haurwitz’s semi-empirical approach, allowing for changes in cloud type and fractional coverage, was developed. The model predicts the annual global radiation fluxes to within 2–11% of the observed values. Estimated monthly mean values gave an average accuracy within about 6% of the measurements. The estimates agree well with the observations during the first four months of the year but less so for the last four. Changing surface albedo might explain this deviation. Previously, the 1993 National Solar Radiation Data Base (NSRDB) from the National Renewable Energy Laboratory (NREL) modeled global radiation data for 16 Alaskan stations. Although more complete and complex, the NREL model requires a larger number of input parameters, which are not available for Alaska. Hence, we believe that our model, which is based on cloud-radiation relationship and is specifically tuned to Alaskan conditions, produces better results for this region. Annual global solar radiation flux measurements are compared with results from global coverage models based on the International Satellite Cloud Climatology Project (ISCCP) data. Contour plots of seasonal and mean annual spatial distribution of global radiation for Alaska are presented and discussed in the context of their climatic and geographic settings. Received July 16, 1997 Revised May 18,1998     

Application of GIS and logistic regression to fossil pollen data in modelling present and past spatial distribution of the Colombian savanna

Climate changes affect the abundance, geographic extent, and floral composition of vegetation, which are reflected in the pollen rain. Sediment cores taken from lakes and peat bogs can be analysed for their pollen content. The fossil pollen records provide information on the temporal changes in climate and palaeo-environments. Although the complexity of the variables influencing vegetation distribution requires a multi-dimensional approach, only a few research projects have used GIS to analyse pollen data. This paper presents a new approach to palynological data analysis by combining GIS and spatial modelling. Eastern Colombia was chosen as a study area owing to the migration of the forest–savanna boundary since the last glacial maximum, and the availability of pollen records. Logistic regression has been used to identify the climatic variables that determine the distribution of savanna and forest in eastern Colombia. These variables were used to create a predictive land-cover model, which was subsequently implemented into a GIS to perform spatial analysis on the results. The palynological data from the study area were incorporated into the GIS. Reconstructed maps of past vegetation distribution by interpolation showed a new approach of regional multi-site data synthesis related to climatic parameters. The logistic regression model resulted in a map with 85.7% predictive accuracy, which is considered useful for the reconstruction of future and past land-cover distributions. The suitability of palynological GIS application depends on the number of pollen sites, the distribution of the pollen sites over the area of interest, and the degree of overlap of the age ranges of the pollen records.     

Modelling the ground heat flux of an urban area using remote sensing data

Summary During the Basel Urban Boundary Layer Experiment (BUBBLE) conducted in 2002, micrometeorological in-situ data were collected for different sites using a variety of instruments. This provides a unique data set for urban climate studies. Nevertheless, the spatial distribution of energy and heat fluxes can only be taken into account with remote sensing methods or numerical models. Therefore, multiple satellite images from different platforms (NOAA-AVHRR, MODIS and LANDSAT ETM+) were acquired, processed and analysed. In addition, a high resolution digital elevation model (DEM) and a 1 m resolution digital surface model (DSM) of a large part of the city of Basel was utilized. This paper focuses on the calculation and modelling of the ground (or storage) heat flux density using remotely sensed data combined with in-situ measurements using three different approaches. First, an empirical regression function was generated to estimate the storage heat flux from NDVI values second approach used the Objective Hysteresis Model (OHM) which is often used for in-situ measurements. The last method used information of the geometric parameters of urban street canyons, computed from the high resolution digital urban surface model. Modelled and measured data are found to be in agreement within ±30 Wm⁻² and result in a coefficient of determination (R²) of 0.95.     

Modeling and mapping temperature and precipitation climate data in Greece using topographical and geographical parameters

This study presents a methodology for modeling and mapping the seasonal and annual air temperature and precipitation climate normals over Greece using several topographical and geographical parameters. Data series of air temperature and precipitation from 84 weather stations distributed evenly over Greece are used along with a set of topographical and geographical parameters extracted with Geographic Information System methods from a digital elevation model (DEM). Normalized difference vegetation index (NDVI) obtained from MODIS Aqua satellite data is also used as a geographical parameter. First, the relation of the two climate elements to the topographical and geographical parameters was investigated based on the Pearson’s correlation coefficient to identify the parameters that mostly affect the spatial variability of air temperature and precipitation over Greece. Then a backward stepwise multiple regression was applied to add topographical and geographical parameters as independent variables into a regression equation and develop linear estimation models for both climate parameters. These models are subjected to residual correction using different local interpolation methods, in an attempt to refine the estimated values. The validity of these models is checked through cross-validation error statistics against an independent test subset of station data. The topographical and geographical parameters used as independent variables in the multiple regression models are mostly those found to be strongly correlated with both climatic variables. Models perform best for annual and spring temperatures and effectively for winter and autumn temperatures. Summer temperature spatial variability is rather poorly simulated by the multiple regression model. On the contrary, best performance is obtained for summer and autumn precipitation while the multiple regression model is not able to simulate effectively the spatial distribution of spring precipitation. Results revealed also a relatively weaker model performance for precipitation than that for air temperature probably due to the highly variable nature of precipitation compared to the relatively low spatial variability of air temperature field. The correction of the developed regression models using residuals improved though not significantly the interpolation accuracy.     

Data-model comparison using fuzzy logic in paleoclimatology

 Until now, most paleoclimate model-data comparisons have been limited to simple statistical evaluation and simple map comparisons. We have applied a new method, based on fuzzy logic, to the comparison of 17 model simulations of the mid-Holocene (6 ka BP) climate with reconstruction of three bioclimatic parameters (mean temperature of the coldest month, MTCO, growing degree-days above 5 °C, GDD5, precipitation minus evapotranspiration, P−E) from pollen and lake-status data over Europe. With this method, no assumption is made about the distribution of the signal and on its error, and both the error bars related to data and to model simulations are taken into account. Data are taken at the drilling sites (not using a gridded interpolation of proxy data) and a varying domain size of comparison enables us to make the best common resolution between observed and simulated maps. For each parameter and each model, we compute a Hagaman distance which gives an objective measure of the goodness of fit between model and data. The results show that there is no systematic order for the three climatic parameters between models. None of the models is able to satisfactorily reproduce the three pollen-derived data. There is larger dispersion in the results for MTCO and P−E than for GDD5. There is also no systematic relationship between model resolution and the Hagaman distance, except for P−E. The more local character of P−E has little chance to be reproduced by a low resolution model, which can explain the inverse relationship between model resolution and Hagaman distance. The results also reveal that most of the models are better at predicting 6 ka climate than the modern climate. Received: 27 May 1998 / Accepted: 8 January 1999     

Spatial analysis of monthly precipitation in Turkey

Summary  The principal objectives of this paper are to develop and validate an optimum interpolation method for the spatial analysis of monthly precipitation in Turkey. A two-dimensional optimum interpolation objective analysis scheme has been developed for the spatial analysis of precipitation. The model is developed for generating statistically optimum interpolation based on the irregular distribution of meteorological stations. One question that affects the optimum interpolation method and, indeed, all such techniques, is how many observations should be allowed to influence a given grid point? The method developed in this paper addresses this question. For the implementation of the method, 52 stations are considered for Turkey, with 30 years of monthly data at each point. It is observed that each monthly average spatial correlation function shows a monotonically decreasing pattern based on 15 km interval averages. The method provides high estimation accuracy in dense station locations such as in northwestern Turkey. Precipitation contour maps obtained by the optimum interpolation method indicate two spatial trends over Turkey which have not been identified in any previous study. Received June 24, 1999/Revised April 26, 2000     

Climate change trend in China, with improved accuracy

We have found that a spatial interpolation of mean annual temperature (MAT) in China can be accomplished using a global ordinary least squares regression model since the relationship between temperature and its environmental determinants is constant. Therefore the estimation of MAT does not very across space and thus exhibits spatial stationarity. The interpolation of mean annual precipitation (MAP), however, is more complex and changes spatially as a function of topographic variation. Therefore, MAP shows spatial non-stationarity and must be estimated with a geographically weighted regression. A statistical transfer function (STF) of MAT was formulated using minimized residuals output from a high accuracy and high speed method for surface modeling (HASM) with an ordinary least squares (OLS) linear equation that uses latitude and elevation as independent variables, abbreviated as HASM-OLS. The STF of MAP under a BOX-COX transformation is derived as a combination of minimized residuals output by HASM with a geographically weighted regression (GWR) using latitude, longitude, elevation, impact coefficient of aspect and sky view factor as independent variables, abbreviated as HASM-GWR-BC. In terms of HASM-OLS and HASM-GWR-BC, MAT had an increasing trend since the 1960s in China, with an especially accelerated increasing trend since 1980. Overall, our data show that MAT has increased by 1.44 °C since the 1960s. The warming rates increase from the south to north in China, except in the Qinghai-Xizang plateau. Specifically, the 2,100 °C?·?d contour line of annual accumulated temperature (AAT) of ≥10 °C shifted northwestward 255 km in the Heilongjiang province since the 1960s. MAP in Qinghai-Xizang plateau and in arid region had a continuously increasing trend. In the other 7 regions of China, MAP shows both increasing and decreasing trends. On average, China became wetter from the 1960s to the 1990s, but drier from the 1990s to 2000s. The Qinghai-Xizang Plateau and Northern China experienced more climatic extremes than Southern China since the 1960s.     

短时间序列气温要素空间插值方法精度的比较研究

利用中国气象局国家气象中心的全国1956年和1998年1月和7月平均气温数据,对综合法与ANUSPLIN软件插值精度进行了比较。结果表明:在具有30 a的月平均气温栅格数据库作为背景的前提下,采用综合法对短时间序列平均气温数据进行空间化处理,其方法简单、方便,误差相对较小,是一种较适合的方法。     

Mapping average annual precipitation in Serbia (1961–1990) by using regression kriging

The appearence of geostatistics and geographical information systems has made it possible to analyze complex spatial patterns of meteorological elements over large areas in the applied climatology. The objective of this study is to use geostatistics to characterize the spatial structure and map the spatial variation of average values of precipitation for a 30-year period in Serbia. New, recently introduced, geostatistical algorithms facilitate utilization of auxiliary variables especially remote sensing data or freely available global datasets. The data from Advanced Spaceborn Thermal Emission and Reflection Radiometer global digital elevation model are incorporated as ancillary variables into spatial prediction of average annual precipitation using geostatistical method known as regression kriging. The R ² value of 0.842 proves high performance result of the prediction of the proposed method.     

Gridded Croatian climatology for 1961–1990

In climatology, one of the most important pieces of information about the climate of a place or a region is information about the Climatological Normals (CLINO)—the average values of meteorological elements for a 30-year period. This kind of information usually comes in tables and is available for different observation sites from national meteorological services or from World Meteorological Organisation publications. The key issue, then, becomes how to interpolate these values over the entire area of interest to get reliable and accurate estimates (maps) of climatic elements. Here, the regression kriging framework has been applied for mapping of 20 climatological parameters for the 1961–1990 period for the 56,594 km² of Croatian territory, with a resolution of 1 km. In total, 152 main and climatological and 567 precipitation-measuring stations have been used in the analysis. Extensive pre-processing of metadata on station co-ordinates has been done, as well as completion of missing monthly averages. The final results are 20 climatological maps available in high resolution together with error maps and accuracy assessment measures.     

Interpolation of climate variables and temperature modeling

Geographic Information Systems (GIS) and modeling are becoming powerful tools in agricultural research and natural resource management. This study proposes an empirical methodology for modeling and mapping of the monthly and annual air temperature using remote sensing and GIS techniques. The study area is Gangetic West Bengal and its neighborhood in the eastern India, where a number of weather systems occur throughout the year. Gangetic West Bengal is a region of strong heterogeneous surface with several weather disturbances. This paper also examines statistical approaches for interpolating climatic data over large regions, providing different interpolation techniques for climate variables' use in agricultural research. Three interpolation approaches, like inverse distance weighted averaging, thin-plate smoothing splines, and co-kriging are evaluated for 4°?×?4° area, covering the eastern part of India. The land use/land cover, soil texture, and digital elevation model are used as the independent variables for temperature modeling. Multiple regression analysis with standard method is used to add dependent variables into regression equation. Prediction of mean temperature for monsoon season is better than winter season. Finally standard deviation errors are evaluated after comparing the predicted temperature and observed temperature of the area. For better improvement, distance from the coastline and seasonal wind pattern are stressed to be included as independent variables.     

基于台站降水资料对不同空间内插方法的比较

如何把离散的气象台站资料通过合适的空间内插方法转变成规则的网格数据,对于气候变化分析和模拟研究具有重要的意义.作者利用中国区域160个常用台站10年降水观测资料为例,分别采用克里格(Kriging)插值、反距离加权、Delaunay三角剖分线性插值、双谐样条(Biharmonic Spline)插值和Cressman客观分析等几种常见的空间内插方法,较为系统地分析和比较了这几种内插方法插值结果之间的异同,对其优缺点和适用范围进行了适当讨论,并采用不同的网格分辨率和影响半径对Cressman客观分析方法做了进一步分析比较.结果表明:在台站分布密集的区域不同空间内插方法之间差异较小,在台站分布稀疏的区域则差异较大;与更加密集的台站观测资料的比较显示,自动调节影响半径的Cressman客观分析方法与双谐样条插值方法误差相对较小.     

On the vertical profile of surface radiative fluxes in southwest Germany

Summary Vertical profile of surface radiative fluxes in an area of heterogeneous terrain in south-west Germany is presented. Main data sets utilized for the study were recorded during the REgio KLIma Projekt (REKLIP). Supporting observational data were provided by the German weather service and German geophysical consultant service. Elevation of the study sites ranged from 212 m a.s.l. to 1489 m a.s.l. From May to September, monthly mean albedo was generally low at the study sites, ranging from 19% to 24%. For the other months, monthly mean albedo lie between 22% and 25% at the lowland site but extended between 27% and 71% at the highly elevated mountain site. Following the altitudinal increase in surface albedo, net radiative flux and radiation efficiency declined with elevation at an annual mean of 1.15 Wm⁻²/100 m and 0.008/100 m respectively. Absorbed shortwave radiation and effective terrestrial radiation showed mean decline of 1.54 Wm⁻²/100 m and 0.34 Wm⁻²/100 m, respectively, with the mean sky-to-earth radiation deficit amounting to about 52 Wm⁻² for the lowland site and 73 Wm⁻² for the highest elevated site. Some empirical models which express shortwave and longwave radiative fluxes in terms of meteorological variables have been validated for the lowland and mountain sites. Monthly mean daily total estimates of solar radiation obtained from ?ngst?m-Prescott relation were quite consistent with observed values. Parameterisation of downward atmospheric radiation under all sky condition was achieved by extending Brutsaert clear sky atmospheric model. Relationship between outgoing longwave radiation and screen temperature at the study sites was best described by an exponential function unlike the linear relationship proposed by Monteith and Unsworth. Net radiative flux for the lowland and mountain sites has been expressed in terms of absorbed shortwave radiation, cloud amount and screen temperature. Received March 5, 2001 Revised October 29, 2001