In this study，Lycium barbarum in Ningxia arid area is taken as study object，the real-time data of daily stem sap flow ( SSF) in the whole growth period ( from April to October) of 2015 was obtained by using the packaged stem sap flow monitoring system，and the soil moisture content in root zone and the meteorological data in experiment zone are obtained through soil moisture monitoring stations and automatic weather stations． By using the meteorological data of Yinchuan meteorological station in recent 10 years ( January of 2005 to December of 2014 ) ，the reference evapotranspiration ( ET0 ) series of Lycium barbarum are obtained，and the simulation model for transpiration water consumption and ET0，and its key elements of the SPAC system is established to provides support for the clarification of the water consumption rules and the water requirement diagnosis of Lycium barbarum． The results showed that: in the growth period，the sap flow ( S) and reference evapotranspiration ( ET0 ) has positively correlation ( P= 0．01) ，fitting the linear model S = 0．330 7 ET0－0．201 ( Ｒ2 = 0．822 1) ; meanwhile，the sap flow and the base diameter ( r) ，root zone soil water content ( θ) ，solar radiation ( Ｒs ) ，Temperature ( T) and saturated vapor pressure difference ( VPD) also show significant positive correlation ( P = 0．01) ，and are in accordance with the multi parameter model S4－10 = 1．378 r+0．017θ+0．004 Ｒs+0．028 T+0．34 VPD－4．579( Ｒ2 = 0．693) ; the error between the simulated values and the measured values is between 5．80% and 2．52%，the cumulative average error is －0．93% ( D= 3．5 cm) and 0．69 ( D= 2．6 cm) ，respectively; the simulation precision is good．

To determine the suitable amount of water application in lettuce under drip irrigation, 4 different treatments of water amount were carried out to study the impact of water amount on height, diameter, vitamin C and nitrate content in lettuce. Result showed that, within the threshold of 4 treatments, different water amount had no significant impact on height, diameter, vitamin C and nitrate content in lettuce, whereas plot spike width, number of leaves, weight of single lettuce and yield in lowest amount of water was significantly lower than in high amount of water. Indexes of lettuce in middle amount of water had no significant difference compared with that in high amount of water, except total soluble solids. Plot spike width, number of leaves, weight of single lettuce in low water amount had no significant difference compared with that in high water amount, whereas total soluble solid was significantly higher than that in lowest water amount and high water amount, and the water used efficiency in low water treatment was the highest, up to 33 kg/m3. Taking plant growth, quality, yield and water use efficiency etc., low water amount(93.3mm) was suitable for lettuce grown in open field in spring, with the yield reaching 2797kg/667m2. Crop coefficients of lettuce in different growing stages were calculated, which could give some instructions to lettuce production in open field in spring under drip irrigation.

A BP-W-T prediction model based on BP neural network is set up，which contained a 17－21－14 topological structure． In the BP neural network model，soil fertilizer rate，irrigation water，the initial temperature of every layer，the layers of the initial moisture content and time are taken as input data，soil temperature and moisture content are taken as output data，Matlab software is used to train the BP neural network and then inspect temperature and moisture content distribution by combining with the measured values． The results show that: the average relative error，correlation coefficient and determination coefficient between simulation value and measured value，which is calculated by soil temperature prediction model，is 2．28% ～ 3．36%，0．954～ 0．972，0．910 2～ 0．944 7，respectively，and is 1．87% ～ 3．09%，0．974 3～ 0．992 6，0．949 3～ 0．985 2，respectively，when soil moisture content prediction model is adopted． The prediction model has higher prediction precision and better stability，can better describe dynamic change of soil water and heat under the temperature gradient，and provides a new way to predict the soil temperature and moisture content．

In order to study the effects of alternate micro-irrigation on the growth of plants，an experiment combined alternate irrigation with micro-irrigation was conducted in greenhouse． Including the conventional irrigation，the experiments set up 3 groups of treatment and repeated two times to study the effects of different water pressures on growth of water spinach． The results showed that different irrigation treatments had significant effect ( P＜ 0．05) on leaf area，plant fresh weight and plant water content，but had no significant effect on plant dry weight ( P＞ 0．05) ; alternating moistube irrigation could obtain higher yield; in the depth of 15 cm and the spacing of 30 cm，the water pressure of 1．0 m was the most favorable for plant growth．

In arid areas with shortage water resources，studying mulch drip irrigation technology to prevent soil salinization has important theoretical and practical significance to ease the tension of water resources and to improve the water resources utilization efficiency． In this paper，the dynamic characteristics of water and salt in spring maize field under different irrigation quota were studied through a two-years field experiment，the yield and water use efficiency of spring maize were also studied． The results showed that when the irrigation quota was 430 m3 /hm2 ，the distribution of soil moisture in the 0～ 100 cm soil layer was uniform and no deep percolation appeared． For the two consecutive years，mulch drip irrigation could ensure that the soil layer was under good water and salt environment，and the surface accumulation of soil salinity would not be caused． The difference of yield and water use efficiency of Spring Maize in the two years was not significant． The T2 treatment was the best irrigation system for the Maize under the condition of mulch drip irrigation．

For the purpose of exploring the effective mode of fertigation，providing quantitative indexes for facilities tomato in intelligent control，an experiment was conducted to analyze the effects of different fertilization on growth，yield and quality of tomato． The plot experiments were conducted in vegetable fertilization with four fertilizer levels，including F1 ( flowering fruit bearing stage N214－P2O5 67－ K2O 254 kg /hm2 and harvest stage N285－P2O5 89－K2O 338 kg /hm2 ) ，F2( flowering fruit bearing stage N107－P2O5 33－K2O 127 kg /hm2 and harvest stage N142－P2O5 44－K2O 169 kg /hm2 ) ，F3( flowering fruit bearing stage N64－P2O5 20－K2O 76 kg /hm2 and harvest stage N85－ P2O5 27－ K2O101 kg /hm2 ) ，F4( no fertilizer) ，and the same irrigation water amount． The results showed that fertilizer rate under the test condition had no effect on tomato leaf number，and had positive correlation with plant height and leaf area，inverse correlation with stem diamete; fertilization rate had negative effect on single fruit weight of autumn-winter crop tomato; tomato yield appeared with“parabola” trend with the increase of fertilizer rate; increasing amount of applied fertilizer made the nitrate accumulation increase; soluble sugar and Vc appeared with“parabola”trend with the increase of fertilization rate． Synthetically considering tomato yield and quality，a water-saving， high-yield，good-quality and high-efficiency fertigation system under the controlling drip irrigation was found with fertilization rate of N107－ P2O5 33－K2O 127 kg /hm2 in flowering fruit bearing stage and N142－P2O5 44－K2O 169 kg /hm2 in harvest stage．

Four spring wheat of plateau448, plateau437, plateau602 and Abbondanza were chosen to conduct field experiments, which designed by L16 (43), researching the differences of the water use efficiency and yield under different water low limit (50%, 60% ,70% of the field capacity and CK）and different fertilizers (0 kg/hm2, 105 kg/hm2, 210 kg/hm2 and 315 kg/hm2) . The results showed that different levels of factors had significant effects on water use efficiency and yield of spring wheat. Plateau 448 and plateau 437 had higher yield than that of plateau 602 and Abbondanza, water level CK’s yield was greater than or equal to the yield of 60% and70%,which yield was higher than that of 50%, the yield of fertilizer level 210kg/hm2 was greater than or equal to the yield of 105 kg/hm2 and 315 kg/hm2, which yield was greater than or equal to that of 0 kg/hm2; Water use efficiencies appeared that plateau 448 greater than or equal to plateau 437, plateau 602 and Abbondanza, water level CK’s WUE was higher than 60%,70% and 50%. Fertilizer had higher WUE than that of no fertilizer. The optimization scheme for highest WUE and yield is plateau 448, soil water of 60% and 70% and 210kg/hm2.

In order to evaluate and study the applicability and application effect of maize straw mulching in Hetao irrigation district，three treatments，including plastic film mulching，straw mulching and non mulching treatment，were used to compare the effects of different mulching methods on soil water-fertilizer-heat condition，yield and water use efficiency of Maize． The results showed that there was no significant difference for 0～ 100 cm depth soil water content of plastic film mulching and straw mulching ( P＞0．05) ，both had a better water retention effect，and both the 0～ 60 cm depth soil water content was significantly higher than that of non mulching treatment( P＜0．05) ; The 0～ 60 cm depth soil water content of straw mulching treatment was significantly higher than that of other two treatments in the later stage of maize growth，which indicated that straw mulching was more conducive to soil water conservation and sustainable utilization; straw mulching could improve soil nutrient content in the surface layer of 0～ 20 cm depth effectively，and the available potassium and available phosphorus were the most significant; straw mulching could regulate the soil temperature effectively，which provided the suitable soil temperature condition for the normal growth of maize; there was no significant difference between straw mulching and plastic film mulching for yield and water use efficiency of maize ( P＞0．05) ，but it was average increased by 24．18% and 23．64% than non mulching treatment and reached significant difference ( P＜0．05) ． The comprehensive analysis shows that maize straw mulching planting mode is more suitable for application in Hetao irrigation district．

Drought and water shortage has become major limiting factors for the development of Chinese agriculture，especially in the arid and semi -arid regions． By selecting hybrid millet with better drought resistant performance and higher yield and combining with the technique of mulching plastic film and decreasing planting density，efficient water use in arid region can be achieved． The study results show that: ① under the condition of only one-time irrigation of 1 200 m3 /hm2 before sowing and almost no rainfall supplement in Dunhuang city，the yields of hybrid millets generally can reach 5 000 kg /hm2 ，and the highest yield can be up to 6 106． 05 kg /hm2 ． ② The Pearson correlation coefficient analysis shows that the correlation coefficient between grain weight per plant and tiller number is 0．575 93，which is moderate relevant and reachs a significant level． ③ The technique of Mulching plastic film and decreasing planting density can effectively reduce the ineffective loss of soil moisture，and solve the incompatible contradiction between water demand for hybrid millet and rainfalls． The study has important significance for effectively dealing with occurrence of extreme weather，protecting water resources，and ensuring food security． 

In order to reveal the effect of brackish water with different salinity on the distribution of moisture content and salinity in the soil and the growth characteristics of summer squash，the greenhouse cultivation experiment was conducted． The moisture content and salinity in soil，the emergence rate，the leaf area index，and the yield were observed regularly． The results showed that: the distribution of moisture content and salinity were significant under different salinity treatment; a high salinity equated to a narrow distribution on radial direction and a deep distribution on vertical direction; the three treatments showed a similar result that the salinity would accumulate in the depth of 0～ 20 cm and desalinize in the depth of 20～ 40 cm; the main area of salinity accumulation was 20～ 30 cm under the treatment of 1．7 g /L，while the result was 30～ 40 cm under the other two treatments． The soil electrical conductivity under different treatments followed a descending order of 1．7，3．5，5．1 g /L; the increase of salinity would devote the decrease of emergence rate and the extension of emergence time，it would also had an inhibiting effect on the leaf area index and yield．

Basing on the data of monthly precipitation，average monthly temperatures，sunshine hours and average relative humidity from 1959 to 2014 of Jiamusi，which is a major grain production region of Sanjiang Plain，the periodic features and factors of precipitation since 1959 are analyzed with the methods of wavelet transformand and multiple linear regression． The results show that the precipitation has three obvious main oscillation periods of 12，288 and 480 months; the main factors are the average relative humidity and the average temperature，followed by the average air pressure and sunshine hours，and the last is the average wind velocity． The research results can provide theoretical basis for efficient utilization of rain resources in major grain production region of Sanjiang Plain

Virtual water strategy is one of the efficient paths of agricultural industrial structure optimization． By calculating and analyzing the virtual water contents of the main crops in Guangxi，the agricultural industrial structure is optimized． The results showed that there is large difference between the virtual water contents of the main crops; the virtual water content of rapeseed is 8．733 m3 /kg，which is 54 times than that of vegetable; meanwhile，there is also large difference in the change trend of the virtual water contents of the main crops during 2000 to 2014; the change trends of the virtual water contents of rice，sugarcane，cassava and vegetable are smooth and steady，those of wheat， rapeseed and tobacco are fluctuant，and those of beans，potato and maize are declining; all the cultivated area of the main crops have varying degrees adjustment after optimized． Compared with the current planting pattern，the production values of the main crops under the economic planting pattern is increased by 5．43%，but the total water consumption is still large． The production values of the main crops under the ecological planting pattern maintain unchanged，but the total water consumption is decreased by 5． 47%． On this basis，the optimizing countermeasure of agricultural industrial structure in Guangxi is proposed．

In an acute shortage of water resource of Hebei province, how to use scarce water resources reasonably has become a hot topic. Being able to protect and take full advantage of these pit-ponds reasonably, plentiful water was stored to actualize using summer water in spring, and being able to reduce the exploitation of underground water, which can make full use of water resources. Based on images of Landsat TM, ETM+ and OLI satellites, by using modified normalized difference water index ?(MNDWI) of remote sensing technology, pit-ponds water in groundwater restricted exploitation area in Hebei province was extracted, thereby doing statistical analysis of number and area of pit-ponds and estimating capacity of pit-ponds water storage in study area. Based on these, distribution of pit-ponds presents a pattern that pit-ponds in north were more than south and ponds in east were more than west. Under the influence of climate and human factors, during 2000 to 2014, the number and area of pit-ponds water in groundwater restricted exploitation Area in Hebei province were declined year by year. Analysis of pits or volatility as irrigation water source protection capacity was also discussed in this paper. Water storage of pit-ponds in 2000, 2007 and 2014 were 5.09×108, 3.76×108 and 1.95×108. This research had great significance to agriculture irrigation in this area.

The risk of water resources scarcity in 31 provinces and autonomous regions of China from 2004 to 2014 is analyzed by using grey relational analysis and improved ant colony clustering algorithm． The results show that: ① Industrial water，ecological water，total water resources and domestic water have the greatest impact on the risk of water scarcity in China． ② For the average value of 20 iterative experiments based on Iris and Zoo data sets，improved ant colony clustering algorithm，compared to basic ant colony clustering algorithm， increases the F-measure by 1．01% and 0．64%，respectively，and decreases the total deviation error by 32．057 and 24．926，respectively． ③ According to the cluster analysis results of water resources scarcity in China，the threshold value of classification is established by normal distribution． The provinces and autonomous regions with the low risk of water resource scarcity are Zhejiang，Jiangxi，Anhui，Fujian，Hubei， Hunan，Guangdong，Guangxi，Sichuan，Guizhou，Yunnan，Tibet，Xinjiang，Qinghai and Heilongjiang; Chongqing has the medium risk and other areas have high risk． For the seven regions，the low risk regions are the center and southwest of China，the medium risk regions are the east south and the northwest of China，and the high risk regions are the north and northeast China．

The research on crop water saving and salt control is of great significance to realize the efficient utilization of water resources and the sustainable development of agriculture． Taking the Kongque Ｒiver Basin in Xinjiang as the study area，the cotton growth characteristics in different degree of soil salinization is studied under two different irrigation quotas． The results show that: under the same irrigation frequency， irrigation quota and plant height has positive correlation; under the same irrigation quota，irrigation frequency and plant height has positive correlation; for high salinity soil，low irrigation quota with high irrigation frequency can obtain more cotton bolls number and the single boll weight is heavier，large irrigation quota with low irrigation frequency can obtain more cotton bolls number but the single boll weight is lighter; for low soil salinization soil，low irrigation quota with low irrigation frequency can obtain more cotton bolls number but the single boll weight is lighter，large irrigation quota with large irrigation frequency can obtain more cotton bolls number but the single boll weight is lighter; high irrigation frequency can enhance cotton yield and water productivity for planting cotton in salinization soil．

The information of wetland cover in Liaohe estuary is extracted by 3S technology to provide the data base for calculation of ecoenvironmental water requirement． According to the actual situation and the characteristics of ecosystem，the calculation types and methods of water requirement of eco-environment is determined． The calculation results show that the optimum water requirement of eco-environment is 2．74×108 m3 /a; the minimum water requirement of eco-environment are 0．78×108 m3 /a，0．94×108 m3 /a and 1．10×108 m3 /a，respectively， under the condition of average annual rainfall，75% probability and 95% probability． In order to realize the optimal allocation of water resources，the progress of eco-environment water requirement month by month is given． The minimum water requirement of eco-environment is essential important in the case of water shortage．

This paper analyze water status and trends of the Manas river basin in 2001-2014 from the department and regional structure, as well as the water efficiency in various areas; We use Support vector machine (SVM) regression method to construct water requirement model, and predict the short-term water demand, then we use principal component analysis (PCA) to analyze the influence factors of water requirement. Water consumption of Manas river basin changed little in the study period, the proportion of agriculture is more than 90% and is falling, but the other departments have a certain rise; Regional water consumption is relatively fixed, Shihezi accounts for the largest proportion, followed by Shawan and Manas; Watershed unit water output and primary industry output are low and grow slowly, the second industry output is the highest and grows rapidly; The regional unit water outputs are similar; River basin water use efficiency is high. Support vector machine (SVM) regression method is good at fitting and forecasting Manas river basin water demand. The regression model predicts that Manas river basin total water use is slightly decreased in the next few years. Population, economic and social developments are the main factors influencing the water demand, next is the climate factors such as temperature and precipitation.

In order to discuss the applicability of biological carbon for maize planting in Hetao irrigation district，five treatments，including non biological carbon( CK) ，biological carbon application amount for 15 t /hm2 ( ST1) ，30 t /hm2 ( ST2) ，45 t /hm2 ( ST3) and 60 t /hm2 ( ST4) ，are set to study the effects of different amount of biological carbon on soil moisture，soil temperature，plant height and yield of maize． The results show that: the soil moisture content of different treatments increases first and then decreased with the increase of biological carbon amount，but all are significantly higher than that of CK，and the effect with applied amount of 45 t /hm2 is the most significant，which is increased by 12．95%，14．83%，11．06%，8．86% and 10．40% than that of the control treatment CK; with the increase of the biological carbon amount，the average temperature of the surface 0 ～ 25 cm soil shows increasing trend，and different biological carbon application amount is significantly higher than that of the control treatment CK during the whole growth period( P＜0．05) ; when the amount of biological carbon is 60 t /hm2 ，the effect of soil temperature increasing is the most significant，and that in the whole growth period is significantly higher than those of other treatments( P＜0．05) ; the appropriate application of biological carbon can significantly promote the growth and development of maize，the maize plant height shows the change law of first increase and then decrease with the increase of biological carbon application amount during the whole growth period; the biological carbon application amount of 45 t /hm2 has the optimal effect for maize plant height growth and development; biological carbon can significantly improve maize yield，and the maize yield of treatment ST1，ST2，ST3 and ST4 isincreased by 8．93%，14．14%，17．09% and 15．43% than that of the control treatment CK，the difference is significant ( P＜ 0．05) ． In summary，biological carbon can significantly improve the hydrothermal environment of soil and promote the growth and yield of maize． According to the comprehensive analysis，the biological carbon application amount of 45 t /hm2 is the most suitable for maize planting in Hetao irrigation district．

Knowledge acquisition has always been a key problem in constructing decision support system，and building a comprehensive and effective knowledge base is of great significance to decision support system． Knowledge of agricultural irrigation is not only the experience of experts or some mathematical models can be summarized，many of the laws of objective existence has not yet been recognized． In this paper， the methods of automatic knowledge acquisition are presented，including knowledge acquisition method based on web pages，API and text， and knowledge base of agricultural irrigation is built to provide decision support for agricultural irrigation． Vision is the main way for people to accept information，and human can quickly understand a lot of visual information． This paper firstly analyzed the relationship between meteorological parameters and crop evapotranspiration by using visual analysis technology in an intuitive and visual way． Secondly，the decision tree was generated by data mining visual analysis technology． The visualization model of decision tree was analyzed and studied，and extracted the rule for updating knowledge base to provide relevant decision-making knowledge for agricultural producers．

As an important precise measuring facility，IC card technology has great significance for promoting the reform of agricultural water price and achieving sustainable development of groundwater resources in rural areas． Based on the field research in Sanhe，Hebei Province， this paper uses the method of ordered logistic model to analyze the influencing factors for households＇ willingness of participation in the IC card system． The survey result shows that 73．46% of the farmers in project villages have the willingness to participate in the IC card system． Households who are males，whose average annual household incomes are higher，who think that the use of IC card has brought benefits ( including irrigation is more convenient and timely，electrical safety of irrigation is increased，water allocation is more fair and reasonable， water tariff collection is more open and transparent，the frequency of irrigation equipment damage is reduced) ，who think that operation is convenient are more willing to participate in the IC card policy． Households whose families have more number of cultivated land are more reluctant to participate in the policy．

As the basic facility of the irrigation management，the research on the hydraulic characteristics of water-measuring devices is the basic work for water measuring in irrigation area． By using FLUENT 6．3 software with VOF method and ＲNG k－ε turbulence model，3D numerical simulation of the arc angle trapezoidal canal without throated measuring flume is conducted to study the velocity distribution，water line，head loss and other hydraulic characteristics under different conditions． Besides，the relevant flow calculation formula is established and the comparative analysis is made with measured canal discharge． The results show that the formula obtained from numerical method can get accurate flow measurement in the range of 4．985 3～ 33．597 6 m3 /s when the contraction ratio is 0．57 ～ 0．74，and can provide base for the decision of irrigation water use plan and the adjustment of water supply and distribution scheme of irrigation canal system．

The drought occurrence has greatly reduced the grain yield of Heilongjiang Province with the commodity grain production base of the largest agricultural acreage in China． However，the research on drought was relatively not enough with the uncertainty of application of drought index． Taking Harbin city as an example，the year distribution of precipitation was analyzed by using the long series precipitation data from 1951 to 2014，and the occurrence of drought at different scales was analyzed by using the precipitation anomaly percentage index． The results showed that this index was suitable for Harbin City，and the severity of drought was presented in the descending order of 1970s，21 century，1960s，1990s，1950s and 1980s． The order of drought occurrence was winter，spring，autumn，summer． The severe，moderate and mild droughts were easy to occur in Harbin city at April，August and November，respectively． While the probability of severe drought was unstable，the mild and moderate droughts had a great probability．

In order to timely and accurately obtain irrigation water demand data，using ArcEngine and C # as a development platform and language，combined with GIS and ＲS data analysis and image processing functions，an application system of calculating water requirement for irrigation district is developed． The system is based on the soil water content of different crops and the land classification images of the irrigated areas by remote sensing thermal infrared band，and the calculation parameters are designed according to the irrigation system of different crops． Through system operation，the irrigation water demand of irrigation area can be obtained quickly． The results of case data of Jinghui irrigation district in Shaanxi Province show that the system has dynamic monitoring and practical effect and can provide important basic data for guiding irrigation area to optimize water resources allocation and irrigation．

Because the hardware and software should be developed step by step in different time，there are several issues in the traditional water and fertilizer irrigation controller development，such as algorithm embedded in complex，long development process and hardware and software debug cycle etc． This paper describes an efficient and novel method of irrigation water and fertilizer controller rapid development based on MBD-TSP( Model Based Designer，Target Support Package) ． The model of water and fertilizer controller is established based on Simulink environment and an embedded peripheral hardware control system is designed based on STM32 chip． After Multi-layer simulation of embedded algorithm，code is produced with Coder instrument． Finally，the HEX file generated by IDE is validated for water and fertilizer irrigation controller． The results show that with the new development control system method，the development cycle is reduced by about 50% than before and the algorithm is verified in advance in different links of the development process． Meanwhile，the experiment of water fertilizer concentration and fertilizer consumption precise control shows that the new algorithm has response speed and good real-time performance and the irrigation effect is greatly enhanced．