基于多元非线性回归与BP神经网络模型对土壤水分蒸发量的预测

为探讨农田小气候中各种因素(温度、湿度、光合有效辐射、环境梯度)对土壤水分蒸发量的影响,以河海大学江宁校区节水园区的壤土作为研究对象,根据2020年9-11月期间每30 min测定一次的数据资料,利用多元非线性回归与BP神经网络算法对其进行预测。结果表明,多元非线性回归与BP神经网络均可以较好地得到预测结果;相比之下,BP神经网络算法预测值的拟合优度更高。

杉木人工林地土壤水分蒸发的时空变化特征

为了研究土壤水分蒸发规律,做好蓄水保墒,减少水分的非生产消耗,需要了解水分运动规律。本实验利用2012年湖南会同定位观测站第Ⅱ集水区实测数据,研究杉木人工林地土壤水分蒸发日变化、月变化、及不同坡位的变化规律。结果表明:土壤水分观测日最大分时蒸发量发生在9:00～11:00时段,月均蒸发量在6月中旬左右达到最大峰值;春季3～5月期间土壤含水量的平均变化程度较小,而夏季6～7月期间土壤含水量的平均变化程度较大;不同坡位因土壤层厚度、空气流动程度及林下植被凋落物的种类和混合比例不同,其土壤水分蒸发量呈现明显差异。结合这些特点,试图分析其形成的原因,为后期研究提供了参考方向。

定量遥感反演作物蒸腾和土壤水分利用率的区域分异

首先讨论了在农业生态站建立的行之有效的、以遥感数据为主体的作物蒸腾二层模型.重点阐明了以多时相辐射温度和热惯量信息为基础的作物地表混合像元（视场）温度的分解模型,在遥感反演过程中提出了一些在模型和算法方面的优化和改进:提出了一个区域比辐射率估算的算法、提出了运用地表温度的静态反馈算法和扩展生态台站的空气温度到区域的算法.并在LUCC分类技术的支持下,提出了以地面粗糙度和辐射温度为参数的地面2m高的气温和风速的空间扩展算法.最后应用 NOAA-AVHRR和地面同步观测及其定标,按像元逐点运算,反演了华北地区农田作物蒸腾和水分利用率的区域分布.揭示了春天华北地区水资源在农业上的利用效率,为采取有效的节水措施提供了科学依据.

生物质炭保水剂的吸水保水性能研究

以生物质炭为原料，选择保水性和黏结性较强的Mw型、阴型和阳型等聚丙烯酰胺（PAM）对生物质炭进行了亲水性改性，分析了改性后的生物质炭保水剂吸水特性的变化以及对土壤水分蒸发和苗期玉米生长的影响。研究结果表明，施用3％生物质炭两周后可降低土壤水分蒸发量4．1％；施用2．5％和5％的生物质炭后盆栽土壤含水量分别比对照显著提高了39．7％和50．4％。通过PAM改性后的生物质炭可显著提高保水能力，施用改性的生物质炭后盆栽土壤含水量较生物质炭本身提高了10．5％～63．0％。3种生物质炭保水剂的吸水倍率均表现为2：10型〉1．5：10型〉1：10型，说明生物质炭保水剂的吸水倍率随PAM的质量比的升高而升高，且Mw型的吸水效果最好。施用2．5％和5％保水剂的土壤含水量分别比对照提高了54．3％～84．0％和119．5％～145．1％，但从土壤保水作用和对作物生长两方面出发，2．5％的生物质炭保水剂对其持水性和玉米植株的生长较为适宜。

Evapotranspiration and Soil Moisture Balance for Vegetative Restoration in a Gully Catchment on the Loess Plateau, China

Evapotranspiration, soil moisture balance and the dynamics in a gully catchment of the Loess Plateau in China were determined with 6 land use treatments including natural grassland, shrubs （Caragana rnicrophylla）, two woodlands （Prunus armeniaca var. ansu and Pinus tabulaeformis）, cultivated fallow, and farmland （Triticum aestiuum L.） in order to obtain a better understanding of soil moisture balance principles and to improve vegetation restoration efficiency for ecological rebuilding on the plateau. Average runoff from cultivated fallow was very high, reaching 10.3% of the seasonal rainfall. Evapotranspiration under T. aestivurn was not significantly different from natural grasslands. Compared with natural grass, evapotranspiration was significantly greater （P 〈 0.05） in 2002 and there was an increase in soil moisture depleted in the 1-3 m soil under P. armeniaca, P. tabulaeformis and C. microphylla. During the two years of the study the average soil moisture （0-100 cm soil profile） of T. aestivurn was generally the highest, with P. armeniaca, P. tabulaeformis and C. rnicrophylla usually the lowest. Thus, according to the soil moisture balance principle for this area the planned reforestation project was not ecologically reasonable. Reducing human disturbance and restoration with grass could be more effective.

Evaporation from Bare Soil in Extremely Arid Environment in Southern Israel

Microlysimeters of different sizes(5 cm 10 cm and 15 cm in length)were used extensively in the present study for the measurements of soil evapondion in site in an extremely arid area in southern Israel.All of the data obtained from the microlysimeters were used to evaluate two conventional eVaporation models developed by Black et al.and Ritchie,respectively.Our results indicated that the models could overestimate total cumulative evaporation by about 30% in the extremely arid environment.Reducing the power factor of the conventional model by a faCtor of 0.1 produced good agreement between the measured and simulated cumulative evaporation.Microlysimeter method proved to be a simple and accurate approach for the evaluation of soil evaporstion.

Evapotranspiration, Yield and Crop Coefficient of Irrigated Maize Under Straw Mulch

Maize (Zea mays L.), a staple crop grown from June to September during the rainy season on the North China Plain,is usually inter-planted in winter wheat (Triticum aestivum L.) fields about one week before harvesting of the winterwheat. In order to improve irrigation efficiency in this region of serious water shortage, field studies in 1999 and 2001, twodry seasons with less than average seasonal rainfall, were conducted with up to five irrigation applications to determineevapotranspiration, calculate the crop coefficient, and optimize the irrigation schedule with maize under mulch, as well asto establish the effects of irrigation timing and the number of applications on grain yield and water use efficiency (WUE)of maize. Results showed that with grain production at about 8 000 kg ha-1 the total evapotranspiration and WUE ofirrigated maize under mulch were about 380-400 mm and 2.0-2.2 kg m-3, respectively. Also in 2001 WUE of maizewith mulch for the treatment with three irrigations was 11.8% better than that without mulch. In the 1999 and 2001seasons, maize yield significantly improved (P = 0.05) with four irrigation applications, however, further increases werenot significant. At the same time there were no significant differences for WUE with two to four irrigation applications.In the 2001 season mulch lead to a decrease of 50 mm in the total soil evaporation, and the maize crop coefficient undermulch varied between 0.3-1.3 with a seasonal average of 1.0.

Responses of soil moisture to vegetation restoration type and slope length on the loess hillslope

Soil moisture, a critical variable in the hydrologic cycle, is highly influenced by vegetation restoration type. However, the relationship between spatial variation of soil moisture, vegetation restoration type and slope length is controversial. Therefore, soil moisture across soil layers（0-400 cm depth） was measured before and after the rainy season in severe drought（2015） and normal hydrological year（2016） in three vegetation restoration areas（artificial forestland, natural forestland and grassland）, on the hillslopes of the Caijiachuan Catchment in the Loess area, China. The results showed that artificial forestland had the lowest soil moisture and most severe water deficit in 100-200 cm soil layers. Water depletion was higher in artificial and natural forestlands than in natural grassland. Moreover, soil moisture in the shallow soil layers（0-100 cm） under the three vegetation restoration types did not significantly vary with slope length, but a significant increase with slope length was observed in deep soil layers（below 100 cm）. In2015, a severe drought hydrological year, higher water depletion was observed at lower slope positions under three vegetation restoration types due to higher transpiration and evapotranspiration and unlikely recharge from upslope runoff. However, in 2016, a normal hydrological year, there was lower water depletion, even infiltration recharge at lower slope positions, indicating receiving a large amount of water from upslope. Vegetation restoration type, precipitation, slope length and soil depth during a rainy season, in descending order of influence, had significant effects on soil moisture. Generally, natural grassland is more beneficial for vegetation restoration than natural and artificial forestlands, and the results can provide useful information for understanding hydrological processes and improving vegetation restoration practices on the Loess Plateau

Daily SPEI Reveals Long-term Change in Drought Characteristics in Southwest China

Drought is the most widespread and insidious natural hazard, presenting serious challenges to ecosystems and human society. The daily Standardized Precipitation Evapotranspiration Index(SPEI) has been developed to identify the regional spatiotemporal characteristics of drought conditions from 1960 to 2016, revealing the variability in drought characteristics across Southwest China. Daily data from142 meteorological stations across the region were used to calculate the daily SPEI at the annual and seasonal time scale. The Mann-Kendall test and the trend statistics were then applied to quantify the significance of drought trends, with the following results. 1) The regionally averaged intensity and duration of all-drought and severe drought showed increasing trends, while the intensity and duration of extreme drought exhibited decreasing trends. 2) Mixed(increasing/decreasing) trends were detected, in terms of intensity and duration, in the three types of drought events. In general, no evidence of significant trends(P < 0.05) was detected in the drought intensity and duration over the last 55 years at the annual timescale. Seasonally, spring was characterized by a severe drought trend for all drought and severe drought conditions, while extreme drought events in spring and summer were very severe. All drought intensities and durations showed an increasing trend across most regions, except in the northwestern parts of Sichuan Province. However, the areal extent of regions suffering increasing trends in severe and extreme drought became relatively smaller. 3) We identified the following drought hotspots: Guangxi Zhuang Autonomous Region from the 1960 s to the 1990 s, respectively. Guangxi Zhuang Autonomous Region and Guizhou Province in the 1970 s and 1980 s, and Yunnan Province in the 2000 s. Finally, this paper can benefit operational drought characterization with a day-to-day drought monitoring index, enabling a more risk-based drought management strategy in the context of global warming.

Drought and Spatiotemporal Variability of Forest Fires Across Mexico

Understanding the spatiotemporal links between drought and forest fire occurrence is crucial for improving decision-making in fire management under current and future climatic conditions. We quantified forest fire activity in Mexico using georeferenced fire records for the period of 2005–2015 and examined its spatial and temporal relationships with a multiscalar drought index, the Standardized Precipitation-Evapotranspiration Index(SPEI). A total of 47975 fire counts were recorded in the 11-year long study period, with the peak in fire frequency occurring in 2011. We identified four fire clusters, i.e., regions where there is a high density of fire records in Mexico using the Getis-Ord G spatial statistic. Then, we examined fire frequency data in the clustered regions and assessed how fire activity related to the SPEI for the entire study period and also for the year 2011. Associations between the SPEI and fire frequency varied across Mexico and fire-SPEI relationships also varied across the months of major fire occurrence and related SPEI temporal scales. In particular, in the two fire clusters located in northern Mexico(Chihuahua, northern Baja California), drier conditions over the previous 5 months triggered fire occurrence. In contrast, we did not observe a significant relationship between drought severity and fire frequency in the central Mexico cluster, which exhibited the highest fire frequency. We also found moderate fire-drought associations in the cluster situated in the tropical southern Chiapas where agriculture activities are the main causes of forest fire occurrence. These results are useful for improving our understanding of the spatiotemporal patterns of fire occurrence as related to drought severity in megadiverse countries hosting many forest types as Mexico.

Validation of Global Evapotranspiration Product(MOD16) Using Flux Tower Data from Panjin Coastal Wetland,Northeast China

Recent advances in remote sensing technology and methods have resulted in the development of an evapotranspiration(ET) product from the Moderate Resolution Imaging Spectrometer(MOD16). The accuracy of this product however has not been tested for coastal wetland ecosystems. The objective of this study therefore is to validate the MOD16 ET product using data from one eddy covariance flux tower situated in the Panjin coastal wetland ecosystem within the Liaohe River Delta, Northeast China. Cumulative ET data over an eight-day period in 2005 from the flux tower was calculated to coincide with the MOD16 products across the same period. Results showed that data from the flux tower were inconsistent with that gained form the MOD16 ET. In general, results from Panjin showed that there was an underestimation of MOD16 ET in the spring and fall, with Biases of -2.27 and -3.53 mm/8 d, respectively(–40.58% and -49.13% of the observed mean). Results for Bias during the summer had a range of 1.77 mm/8 d(7.82% of the observed mean), indicating an overestimation of MOD16 ET. According to the RMSE, summer(6.14 mm/8 d) achieved the lowest value, indicating low accuracy of the MOD16 ET product. However, RMSE(2.09 mm/8 d) in spring was the same as that in the fall. Relationship between ET and its relevant meteorological parameters were analyzed. Results indicated a very good relationship between surface air temperature and ET. Meanwhile a significant relationship between wind speed and ET also existed. The inconsistent comparison of MOD16 and flux tower-based ET are mainly attributed to the parameterization of the Penman-Monteith model, flux tower measurement errors, and flux tower footprint vs. MODIS pixels.

Evapotranspiration and humidity variations in response to land cover conversions in the Three Gorges Reservoir Region

A new land cover classification system was established for the Three Gorges Reservoir Region(TGRR) after considering the continuity of inundation and the natural characteristics of land cover. The potential evapotranspiration(PET) was predicted using a modified Penman-Monteith(P-M) model. The region's ratio of precipitation to evapotranspiration was calculated as the humidity index(HI). The data obtained was used to analyze climatic responses to land cover conversions from the perspectives of evapotranspiration and humidity variations. The results show that, from 1997 to 2009, the average annual PET increased in the early years and decreased later. In terms of overall spatial distribution, a significant reciprocal relationship appeared between annual PET and annual HI. In 1997,the annual PET was higher in the lower reaches than in the upper reaches of the TGRR, but the areas with high PET shifted substantially westward by 2003. The annual PET continued to increase in 2006, but the areas with high PET shrank by 2009. In contrast, the annual HI showed varying degrees of localized spatial variability. Over the three periods, the dominantforms of land cover conversions occurred from evergreen cover to seasonal green cover, from seasonal green cover to evergreen cover, and from seasonal green cover to seasonally inundated areas, respectively. These accounted for 48.0%, 38.4%, and 23.8% of the total areas of converted land covers in the three periods, respectively. During the period between 1997 and 2003, the main forms of land cover conversions resulted in both positive and negative growths in the average annual PET, while all of them pushed down the average annual HI. From 2003 to 2006, the reservoir region experienced neither a decrease in the annual PET nor an increase in the annual HI. The period between 2006 and 2009 saw a consistent downward trend in the annual PET and a consistent upward trend in the annual HI.

Study on Rational Irrigation for Wheat Production in Fengqiu Region: Ⅰ. Analysis of the Relation Between Soil Water Supply and Wheat Yield

A 5-year experiment on water balance was conducted in a flat rainfed wheat field with an area of 66 m×100 m in Fengqiu, Henan Province, China. Results of the experiment showed that the correlation between wheat yield and water consumption was not significant, but that between wheat yield and the ratio of water supply to Penman evaporation was significant, following a parabolic curve. The water consumption process,as well as the growing season of wheat plant, could be divided into three periods. The first (154 days) was vegetative growth period, during which the water consumption accounted for 35% of the total; the second (65 days) reproductive growth period, during which the water consumption occupied 57%, indicating the importance of water consumption in this period; and the third (5~9 days) maturation period, during which water supply was not important to yield formation. According to the statistics of meteorological data over the years in this region, the hydrological conditions of the five seasons covered a probability range of 74.8%.The results (3.46~5.63 t ha-1) indicated that the productivity of the wheat field under rainfed conditions in this region had a degree of confidence of 74.8%.

In-Situ Evaluation of Evaporation in Layered Soils （Tukulu, Sepane and Swartland）

Knowledge about the influence of soil layers on evaporation is essential for the optimization of infield rainwater harvesting （IRWH） in the semi-arid areas of the Free State province of South Africa. Among the soils earmarked for 1RWH development include the Tukulu, Sepane and Swartland soil types that have contrasting soil layers. These soils have to capture and store rainwater within the soil profile layers away from the evaporation zone. To determine how the three soils release and deliver soil water at the evaporating site, a 21-day evaporation experiment was conducted on pre-drained monoliths. Instantaneous soil water content （SWC） from in-situ and soil water characteristic curve （SWCC） from laboratory was measured. Separate soil samples of 15 mm thickness were also evaporated under the same conditions to establish the extent of drying and hydraulic gradient at the soil surface. The Darcian evaporative flux and unsaturated hydraulic conductivity （K-coefficient） were also determined. At the surface suctions of magnitude greater than 1,500 kPa were observed from all monoliths. Total contributions to evaporation from the Tukulu, Sepane and Swartland were 43, 51 and 70 mm, respectively. The low contributions were explained by the presence of the prismacutanic C-horizon in the Tukulu and Sepane at respective depths of 600 and 700 mm. This layer was associated with the steepest suction gradient that restrained further upward fluxes by subsequent lowering for the K-coefficient with more than two orders of magnitudes within a narrow range of SWC. However, the presence of the pedocutanic B-horizon at depths of 300 mm undermined this restrictive function through the appreciable capillary activity demonstrated by clays at near evaporating surfaces. The shallowness and deficiency in structure of the Swartland was consistent with the high contribution to evaporation that gave this soil a dry soil water regime. It was therefore concluded that the Tukulu offered soil profile layers that could reasonably satisfy the soil water conservation requirements for IRWH.

Study of the temporal and spatial patterns of drought in the Yellow River basin based on SPEI

Drought is one of the severe natural disasters to impact human society and occurs widely and frequently in China,causing considerable damage to the living environment of humans.The Yellow River basin(YRB)of China shows great vulnerability to drought in the major basins;thus,drought monitoring in the YRB is particularly important.Based on monthly data of 124 meteorological stations from 1961 to 2015,the Standardized Precipitation Evapotranspiration Index(SPEI)was used to explore the temporal and spatial patterns of drought in the YRB.The periods and trends of drought were identified by Extreme-point Symmetric Mode Decomposition(ESMD),and the research stages were determined by Bernaola-Galvan Segmentation Algorithm(BGSA).The annual and seasonal variation,frequency and intensity of drought were studied in the YRB.The results indicated that(1)for the past 55 years,the drought in the YRB has increased significantly with a tendency rate of-0.148(10 a)^(-1),in which the area Lanzhou to Hekou was the most vulnerable affected(-0.214(10 a)^(-1));(2)the drought periods(2.9,5,10.2 and 18.3 years)and stages(1961–1996,1997–2002 and 2003–2015)were characterized and detected by ESMD and BGSA;(3)the sequence of drought frequency was summer,spring,autumn and winter with mean values of 71.0%,47.2%,10.2%and 6.9%,respectively;and(4)the sequence of drought intensity was summer,spring,winter and autumn with mean values of 0.93,0.40,0.05 and 0.04,respectively.

Impacts of hydraulic redistribution on eco-hydrological cycles： A case study over the Amazon basin

Hydraulic redistribution(HR)refers to the process of soil water transport through the low-resistance pathway provided by plant roots.It has been observed in field studies and proposed to be one of the processes that enable plants to resist water limitations.However,most land-surface models(LSMs)currently do not include this underground root process.In this study,a HR scheme was incorporated into the Community Land Model version 4.5(CLM4.5)to investigate the effect of HR on the eco-hydrological cycle.Two paired numerical simulations(with and without the new HR scheme)were conducted for the Tapajos National Forest km83(BRSa3)site and the Amazon.Simulations for the BRSa3 site in the Amazon showed that HR during the wet season was small,<0.1 mm day^(–1),transferring water from shallow wet layers to deep dry layers at night;however,HR in the dry season was more obvious,up to 0.3 mm day^(–1),transferring water from deep wet layers to shallow dry layers at night.By incorporating HR into CLM4.5,the new model increased gross primary production(GPP)and evapotranspiration(ET)by 10%and 15%,respectively,at the BRSa3 site,partly overcoming the underestimation.For the Amazon,regional analysis also revealed that vegetation responses(including GPP and ET)to seasonal drought and the severe drought of 2005 were better captured with the HR scheme incorporated.

Influence of land evapotranspiration on climate variations

A coupled numerical model of the global atmosphere with a qualified biosphere (GOALS/LASG) has been used to assess the nature of the physical mechanisms for land-atmosphere interactions, and the impacts of the Asian/North American land-surface evapotranspiration on the regional and global climate. This sensitivity study suggests that the simulated climate would be relatively sensitive to land surface evapotranspiration, especially over the Asian regions. The removal of evapotranspiration in Asia would create a warmer and drier climate to a certain degree. Furthermore, the surface evapotranspiration anomalies would make a substantial contribution to the formation and variation of subtropical anticyclones through the changes in monsoon precipitation and the β -effect, but also make a large contribution to the variations of the atmospheric circulation in the Northern Hemisphere and even the globe. Therefore, besides the traditional perception that we have generally emphasized on the influence of subtropical anticyclones activities on the boreal summer precipitation over the regions of eastern China, the surface evapotranspiration anomalies, however, also have substantial impacts on the subtropical anticyclones through the changes in monsoon precipitation. For this reason, the variation in the internal heating sources of the atmosphere caused by the land surface evapotranspiration and the vapor phase change during the boreal summer is an important external factor forcing the weather and climate.

Spatiotemporal variations of aridity in China during 1961–2015: decomposition and attribution

Changes in global climate intensify the hydrological cycle, directly influence precipitation, evaporation, runoff, and cause the re-distribution of water resources in time and space. The aridity index （AI）, defined as the ratio of annual precipitation to annual potential evapotranspiration, is a widely used numerical indicator to quantify the degree of dryness at a given location. This study examined the effects of climate change on Al in China during 1961-2015. The results showed that the nationally averaged AI experienced a notable interdecadal transition in 1993, characterized by increasing AI （wetter） between 1961 and 1993, and decreasing AI （drier） after 1993. Overall, the decreased solar radiation （solar dimming） was the main factor affected the nationally averaged AI during 1961-1993, while the relative humidity dominated the variations of nationally averaged AI during 1993-2015. However, the roles of individual factors on the changes in AI vary in different subregions. Precipitation is one of the important contributing factors for the changes orAl in almost all subregions, except the Mid-Lower Yangtze and Huaihe basins. Solar radiation has been significantly decreased during 1961-1993 in South China, Southwest China, Mid-Lower Yangtze and Huaihe basins, and the Tibetan Plateau. Therefore, it dominated the trends of AI in these subregions. The relative humidity mainly affected the Mid-Lower Yangtze and Huaihe basins, Southwest China, and the Tibetan Plateau during 1993-2015, hence dominated the trends of Al in these subregions. The changes of temperature and wind speed, however, played a relatively weak role in the variations of AI.

Influences of environmental changes on water storage variations in Central Asia

The spatio-temporal pattern of the global water resource has significantly changed with climate change and intensified human activities. The regional economy and ecological environment are highly affected by terrestrial water storage(TWS), especially in arid areas. To investigate the response relationships between TWS and changing environments(climate change and human activities) in Central Asia, we used the Gravity Recovery and Climate Experiment(GRACE) data, Climatic Research Unit(CRU) climate data and Moderate Resolution Imaging Spectroradiometer(MODIS) remote sensing data products(MOD16A2, MOD13A3 and MCD12Q1) from 2003 to 2013, as well as the slope and Pearson correlation analysis methods. Results indicate that:(1) TWS in about 77% of the study area decreased from 2003 to 2013. The total change volume of TWS is about 2915.6 × 108 m^3. The areas of decreased TWS are mainly distributed in the middle of Central Asia, while the areas of increased TWS are concentrated in the middle-altitude regions of the Kazakhstan hills and Tarim Basin.(2) TWS in about 5.91% of areas, mainly distributed in the mountain and piedmont zones, is significantly positively correlated with precipitation, while only 3.78% of areas show significant correlation between TWS and temperature. If the response time was delayed by three months, there would be a very good correlation between temperature and TWS.(3) There is a significantly positive relationship between TWS and Normalized Difference Vegetation Index(NDVI) in 13.35% of the study area.(4) The area of significantly positive correlation between TWS and evapotranspiration is about 31.87%, mainly situated in mountainous areas and northwestern Kazakhstan. The reduction of regional TWS is related to precipitation more than evaporation. Increasing farmland area may explain why some areas show increasing precipitation and decreasing evapotranspiration.(5) The influences of land use on TWS are still not very clear. This study could provide scientific data useful for the estimation of changes in TWS with climate change and human activities.

An improved technique for global daily sunshine duration estimation using satellite imagery

This paper presems an improved model for global sunshine duration estimation. The methodology incorporates geostationary satellite images by including snow cover information, sun and satellite angles and a trend correction factor for seasons, for the determination of cloud cover index. The effectiveness of the proposed methodology has been tested using Meteosat geostationary satellite images in the visible band with a temporal resolution of 1 h and spatial resolution of 2.5 km×2.5 km, for the Brue Catchment in the southwest of England. Validation results show a significant improvement in the estimation of global sunshine duration by the proposed method as compared to its predecessor （R2 is improved from 0.68 to 0.83, root mean squared error （RMSE） from 2.37 h/d to 1.19 h/d and the mean biased error （MBE） from 0.21 h/d to 0.08 h/d）. Further studies are needed to test this method in other parts of the world with different climate and geographical conditions.