Determining the irrigation demands of rice is the scientific basis for rational allocation and effective utilization of water resources. In this study， 12 typical stations in China were selected， and the field water balance model was constructed to derive the traditional irrigation scheduling during 1960-2019. The Mann-Kendall test was used to analyze the change trend of water balance factors， and the Pearson correlation coefficient and the maximum information coefficient (MIC) were used to analyze the data associations. The results showed that the ET_c， irrigation demand and rainfall utilization rate of rice were in the order of middle-rice (556.39 mm) > late-rice (491.47 mm) > early-rice (393.54 mm)； middle-rice (435.94 mm) > late-rice (346.87 mm) > early-rice (147.52 mm)； late-rice (69.60%) > middle-rice (65.46%) > early-rice (54.68%)， respectively. In the past 60 years， the ET_c of rice mainly showed a downward trend. The irrigation demands of rice in central and northeast China showed an obvious downward trend， while the southwest and north areas showed an upward trend. The influence of rainfall on irrigation demands was mainly dominated by the accumulated rainfall during the rice growth period in northern rice area， while it was mainly dominated by rainfall distribution in southern rice area with abundant rainfall.

In order to alleviate the contradiction between water shortage and high water consumption of wheat in Guanzhong area of Shaanxi province， in this study， a two-year field experiment was carried out to investigate the effect of straw returning and mulching on soil nutrients and water-fertilizer use efficiency of wheat. Four different planting modes were set up， including ridge and furrow mulching with straw returning (FMSR)， ridge and furrow no mulching with straw returning (FNSR)， ridge and furrow mulching without straw returning (FMSN)， and ridge and furrow no mulching without straw returning as a control ( CK ). The results showed that compared with CK， soil water content， soil nutrients，stem dry biomass of wheat， water use efficiency and partial factor productivity of nitrogen for FMSR were increased by 2.73%~29.01%，38.41%~40.23%， 30.6%， 24.49% and 21.7%， respectively. Therefore， straw returning and mulching could effectively promote wheat growth and increase yield under ridge and furrow planting pattern. Considering water saving and yield increasing， irrigation of 1 800 m³/hm² and nitrogen fertilizer of 150 kg/hm² are recommended for wheat planting under the ridge and furrow mulching with straw returning.

Open channel roughness is an important parameter affecting the water transmission and distribution and drainage process in the irrigation district. Obtaining accurate roughness value is an important premise to improve the rationality of open channel design， realize the high-precision simulation of multiprocess hydrodynamic， and improve the quality of water management in the irrigation district. In order to improve the roughness calculation quality of open channel in irrigation district， this paper traces back the development history of roughness characterization from a quantitative perspective， expounds the sources of various roughness calculation formulas and their logical relationship， and analyzes the engineering and physical influence factors， including channel section size， wall roughness， hydraulic structure form， channel smoothness and channel bottom gradient， etc. The physical elements mainly include Froude number and turbulence state. On this basis， firstly， the prototype observation method of roughness is reviewed， because it is the basis of other roughness acquisition methods. Then， the research work such as system identification method， roughness optimization method based on inverse calculation of hydrodynamic equation and roughness synchronous solution method based on data assimilation are described， and the GPU parallel computing technology is discussed. The development trend of roughness calculation method under the background of the rapid development of big data analysis technology and web technology is also discussed. Under the background of massive concurrent threads of GPU， this paper focuses on the combination method between the powerful graphic image recognition ability of artificial intelligence technology and the basic waveform of channel hydrodynamic process and its nonlinear superimposed water wave waveform， as well as the geometric information of hydraulic structures， It is pointed out that this method has the remarkable advantage of obtaining different spatial average roughness in real time， and has great theoretical and application potential.

Goji berries (Lycium barbarum L.) widely planted in arid to semi-arid regions are a functional resource characterized by a homology of medicine and food. This study explored the effects of different fertilizer combinations on the soil activated carbon pool and the microbial community structure in the L. barbarum soil under drip fertigation in Ningxia, northwestern China. A two-year field experiment was conducted using different levels of drip nitrogen (40, 60, and 80 mg/L) and phosphorus (10, 20, and 30 mg/L) fertigation. The results showed that the highest average yield was obtained under treatment with 60 mg/L nitrogen plus 30 mg/L phosphorus. The best microbial community structure diversity and abundance of rhizosphere soil was got under treatment with 60 mg/L nitrogen plus 10 mg/L phosphorus. There were differences in the relative abundance of bacterial communities in rhizosphere soil under different treatments. The dominant taxonomic groups at phylum level were Proteobacteria, Actinobacteria, Chloroflexi, Gemmatimonadetes and Firmicutes. RDA and Heatmap analysis showed that the diversity and composition of bacterial community in rhizosphere soil under different treatments were closely related to soil total organic carbon (TOC), easily oxidized organic carbon (EOC), water-soluble organic carbon (DOC), microbial biomass carbon (MBC) and nitrate nitrogen (NO₃-N). In conclusion, rational fertilizer combinations under drip fertigation can change soil microbial community structure and improve soil fertility effectively, which provide theoretical basis for nutrient efficient utilization of L. barbarum and resource sustainable development in arid areas.

In order to explore how to improve the utilization rate of water and fertilizer under drip irrigation for maize, field irrigation experiments were carried out for two consecutive years (2019-2020). Five treatments (T1 (CK), T2, T3, T4 and T5) with different irrigation quota and organic fertilizer usage were set up to study the effects of different soil water and fertilizer regulation. The results showed that:①different treatments had different responses to maize yield. From 2019 to 2020, the per hectare yield of T5 treatment was the largest of 6 860.57 kg/hm². The yield of T2 and T3，T4 increased by 6.77% and 14.8% and 17.97% compared with that of CK, indicating a significant difference. ② There was a linear fitting relationship between the yield increase rate and CK under different chemical fertilizer usage, indicating that chemical fertilizer could help yield increase within the reasonable range; There was a negative exponential relationship between yield increase rate and CK fitting under different irrigation systems, and the larger the irrigation amount was, the better it was.③Considering the agronomic traits, economic traits of ear, crop yield and yield increase, the optimal irrigation quota of seed corn planting area in Hexi Corridor was 5 400 m³/hm². The irrigation quota of each growth stage was 825 m³/hm² at jointing stage, 825 m³/hm² at big trumpet opening stage, 975 m³/hm² at tasseling stage, 975 m³/hm² at flowering stage, 975 m³/hm² at filling stage and 825 m³/hm² at milk ripening stage. To sum up, water and fertilizer are two factors which are closely related and restrict each other in the growth and development of maize. The irrigation system and fertilization scheme are optimal and feasible, which can provide theoretical and technical reference for efficient water-saving cultivation and industrialization development of seed maize in Oasis irrigated area of Hexi Corridor, realize large-scale promotion and replication of farmers, improve maize yield and drive farmers out of poverty.

The research on the optimization of water distribution in irrigation district canal system is of great significance to the transformation of irrigation district from water management to water demand management.At present， the beetle antennae search algorithm has been widely used in production practice， but it has been less applied in the optimization of canal water distribution.In order to explore the feasibility and rationality of the beetle antennae search algorithm in the field of canal system optimization water distribution， this study took the leakage amount during the water delivery process of the two-stage canal system of the West No. 1 trunk canal in the Dagong irrigation district as the optimization goal， and the channel running time and the water delivery flow as the decision-making variables.The canal system water distribution model was constructed， and the beetle antennae search algorithm was used to optimize the model.The results show that the water distribution time in the results of the beetle antennae search algorithm is 7 days shorter than the traditional water distribution time， and the water distribution efficiency is increased by 35%.The leakage of the canal system is 349 900 m³， accounting for 6.69% of the total water distribution， and the water utilization coefficient of the two-stage canal system has increased to 0.704.The beetle antennae search algorithm performs well in solving the problem of optimization of water distribution in high-dimension canal system， and can be close to the actual operation of the canal system and rationally allocate water resources.

In order to explore the effect of water saving and emission reduction under reclaimed water irrigation of rice， an experimental study on reclaimed water irrigation in paddy plots was conducted at the National Agricultural Environmental Dali Observation and Experimental Station. The experiment had two irrigation modes： flooding irrigation (FI) and alternate wetting and drying (AWD)， and three fertilization regimes： F1 (full growth stage fresh water irrigation with all fertilizers)， F2 (tillering stage， jointing and booting stage reclaimed water irrigation with partial fertilizer application)， and F3 (returning green stage， tillering stage， jointing and booting stage reclaimed water irrigation with partial fertilizer application). The concentration of nitrogen， phosphorus， and chemical oxygen demand (COD) in surface water and ground water at different depths as well as their loss load， and rice nitrogen utilization efficiency and fertilizer replacement rate in the paddy field under reclaimed water irrigation were analyzed. In addition， the nitrogen and phosphorus absorption capacity of paddy fields under reclaimed water irrigation， clean water and fertilizer replacement rate were also studied. The results showed that， the peak value of nitrogen concentration in field surface water under reclaimed water irrigation was small and frequent. The average load of nitrogen and phosphorus loss in runoff was 2.65 and 0.62 kg/hm²， respectively， which increased by 26% and 28.6% compared with clean water irrigation. The concentration of nitrogen and phosphorus in the groundwater of paddy fields increased with the fertilization and reclaimed water irrigation. The concentration of nitrogen and phosphorus generally decreased with the increase in depth. However， the leaching loss under reclaimed water irrigation was 11% less than that under clear water irrigation， which could improve the nitrogen use efficiency of paddy field. In AWD mode， TN， TP runoff and leaching load were all reduced compared with FI. The absorption capacity of nitrogen and phosphorus in reclaimed water were 92%， 81%. The removal rate of COD reached 78.2% after 4-5 days of reclaimed water irrigation. The use of reclaimed water for irrigation will hardly have a negative impact on the environment. The replacement efficiency of reclaimed water irrigation for clean water irrigation can reach about 75%. In FI mode， recycled water brought more fertilizer than that in AWD， the annual nitrogen replacement efficiency of chemical fertilizers at the 75% level year was 35.8%， while the amount and replacement rate of phosphorus were both small.

In order to achieve the goals of low energy consumption， saving investment and high irrigation uniformity of the tree-shaped irrigation pipe network， in this paper， a nonlinear mathematical model of tree-shaped irrigation pipe network is constructed， and a hybrid algorithm of simulated annealing and cuckoo (SACS) was designed by setting pressure， pipe diameter and flow rate as constraint conditions. Combined with the data in literature [14]， four optimization algorithms including genetic (GA) algorithm， Simulated annealing genetic (SAGA) hybrid algorithm， cuckoo search (CS) algorithm and SACS hybrid algorithm were used to solve the model. The optimization results showed that the use of the SACS hybrid algorithm reduced the length of the irrigation pipe by 16.27% and the investment by 2.13% compared with the GA algorithm. Compared with the SAGA hybrid algorithm and the CS algorithm， although the irrigation pipe length was the same， the pipe diameter was greatly optimized and the investment was saved by 3.76% and 17.28% respectively， which verified the effectiveness of SACS hybrid algorithm. Finally， taking the sprinkler irrigation project in the 216 hm² modern water conservancy pilot area in Longtang Datun Tea Park， Shiqian County， Guizhou Province， as an example， the SACS hybrid algorithm was used to optimize the sprinkler irrigation project pipe network. Compared with the traditional design method， the irrigation pipe length was reduced by 12.08 m， and the investment was saved by 17.93%. The optimization effect was obvious. This study provides a new idea for optimal design of tree-shaped irrigation pipe network.

In order to study the flow characteristics of flat gate in small and medium-sized rectangular channel under free outflow， this paper used the channel hydraulics test platform to study the water discharge performance of channel flat gate. In this study， the gate opening， upstream water level and flow were tested， and the gate passing flow was calculated by using Garbrecht formula， Duyu formula and Wuhan Institute of Water Resources formula. Compared with the measured flow and other data， it was found that the deviation of the three was large. In order to further improve the calculation accuracy of sluice discharge， the least square method was used to fit the relation between measured discharge coefficient {\mu }_{\mathrm{0}} with relative opening degree of gate， e / H_{\mathrm{0}}， and then a new calculation formula of flow coefficient was obtained. By comparing and analyzing the calculation results of the formula with Garbrecht formula， Duyu formula and Wuhan Institute of water conservancy formula， it was found that the maximum relative error and average relative error of the formula were reduced by more than 16.49% and 17.23%， respectively， and the variation coefficient was also at a small value and within the allowable range of the standard. The results show that for this type of flat gate， the quadratic polynomial formula fitted in this paper has higher accuracy and better applicability， and can meet the measurement accuracy requirements of irrigation area.

In order to improve the level of automation and intelligence of irrigation systems， an intelligent irrigation control system (IICS) based on the embedded system of ARM and the power line carrier was designed. IICS consisted of five modules： data processing and control module；data communication module；data acquisition module；control driver module；human-computer interaction module. The CPU module adopted a 32-bit microprocessor STM32F103CBT6 based on ARM Cortex-M3 architecture. The communication module of the power line carrier adopted a bus master station controller， PB620 chip. In addition， the software of IICS adopted the real-time operating system， μ C/OS-II， with kernel version V2.91. Both the prediction model of soil water shortage and the estimation model of irrigation amount were constructed based on the real-time soil moisture data and short-term weather forecasts. The result showed that the system has realized the functions of soil moisture monitoring， intelligent calculation of irrigation amount and automatic rotation irrigation. The power line carrier realized the power supply and communication functions of soil moisture sensor and solenoid valve， and saved the communication cable. The average packet loss rate of network communication was 0.09%， the bit error rate of the power line carrier was less than 0.01%， and the average response time of the solenoid valve was 0.497 s. Compared with the experience-based irrigation scheme， the model-generated irrigation scheme saved water by 31.37% under the same grain yield level. The system has the characteristics of simple operation， low installation cost， reliable and stable operation， automatic estimation and adjustment of irrigation amount， which can effectively improve the irrigation efficiency and water efficiency， and has a good application prospect.

In order to explore the effects of different moisture content， zeolite content and burial depth on the photosynthetic characteristics of tomato， "aoguan 8" was used as the test material， and the water content（W_50-70， W_60-80， W_70-90）， zeolite content (Z₃， Z₆， Z₉) and burial depth (H₁₅， H₃₀， H₄₅) were used as the experimental factors. The net photosynthetic rate (Pn)， stomatal conductance (Gs) and transpiration rate (Tr) of Tomato during fruit expansion were studied. The results showed that water content was positively correlated with Pn， Gn and Tr. The amount of zeolite was negatively correlated with Gs， which promoted and then inhibited Pn and Tr respectively. There was a negative correlation between burial depth and Pn， and the effects on Gs and Tr were first inhibited and then promoted. In addition， water had a significant effect on Pn (P < 0.05) and Gs (P < 0.01)， but had no significant effect on Tr. The amount and depth of zeolite had no significant effect on the indexes. The order of the influence of the three factors on Pn and Gs was W > Z > H， and the order of influence on Tr was W> H > Z. W_70-90Z₆H₁₅ was the best combination for photosynthetic characteristics of tomato.

In order to reveal the impact of the interaction of the first drought and the subsequent flood on the growth and yield of soybeans under the condition of drought-flood abrupt alternation， a comparative experiment with different degrees of single flooding and different degrees of drought-flood abrupt alternation was carried out. The changes of soybean plant height and chlorophyll content and the law of yield reduction under different treatments were analyzed， the effect of the interaction between drought and flood was quantified， and the law of soybean yield reduction under the drought-flood abrupt alternation was initially revealed. The results showed that the stress effect of drought and flood on the normal growth and development of soybeans and the resulting reduction in yield were significantly less than that of the single flood group； the earlier mild drought had different degrees of mitigation and compensation effects on the suppression of the later flood stress on the normal growth of soybeans. Within a certain range of drought degree， it basically showed that the longer the drought in the early period， the greater the compensation effect in the later period； the reason for the obvious increase in output of the drought-flood abrupt alternation group compared with the single-water logging group was due to the suitable lightness in the early period. The degree of drought compensated 100-seed weight of soybeans by 4.79%~20.00%， the average number of pods per plant by 4.28%~46.63%， and the number of real pods per plant by 9.72%~73.04%. Combined with changes in plant height and chlorophyll content， the law showed that the short-term mild drought enhanced the tolerance of soybeans to the external water and soil environment， and alleviated the physiological stress caused by the later waterlogging on the growth and development of soybeans. The research results can provide technical support for exploring the mechanism of the drought-flood abrupt alternation， and formulating reasonable disaster mitigation and prevention measures for the drought-flood abrupt alternation.

Super absorbent polymers (SAP for short) is a kind of water absorption and water retention ability of high polymer. SAP is widely used in agriculture. This is mainly due to the fact that SAP can maintain and improve the availability of soil moisture and nutrients， reduce soil evaporation， reduce water infiltration， and leachate of fertilizers and heavy metals on the soil profile. At the same time， SAP can also reduce pesticide residues， improve soil microbial activity， increase soil particle agglomeration， improve soil structure. In addition， it can reduce the sensitivity of crops to soil salt， improve the ability of drought resistance， saline-alkali resistance and water and fertilizer utilization. Therefore， it can promote crop growth and increase yield， especially in areas with poor water and fertilizer conditions， where the application effect is more obvious and the benefit is significant. The performance of different types of SAP varies greatly. Its influence on soil and crops is related to many factors， such as raw materials， particle size， pH value of swelling medium， soaking time， medium concentration， temperature， charge number and ionic strength， application method， application amount and soil water content. And there is interaction among the factors. The law is complex and the research results are quite different. This paper mainly discussed the main types of water retaining agent， their water absorption and release performance， and the influence of water retaining agent on soil and crops. At the same time， the water absorption and release mechanism of sap， the main factors affecting water absorption and release， the effect of SAP on soil physical properties and hydraulic properties， and the effect and mechanism of SAP on crop emergence and yield were analyzed. According to the existing main problems， the future research direction was put forward in order to provide reference for SAP to better serve agricultural production.

In order to study the influence of removing the soil background on the unmanned aerial vehicles（UAV）multi-spectral monitoring of soil moisture content，the multi-spectral image data was obtained through UAV flight shooting， the soil background was removed by supervised classification， and the spectral reflectance of 6 bands before and after the background removal were extracted. The reflectance and the measured water content at depths of 10 cm，20 cm and 30 cm were used to construct univariate linear regression model， stepwise regression model， apartial least- squares regression model， and ridge regression model， respectively. R²， RMSE， and RE were used as indicators to evaluate the accuracy of the model. The results showed that the correlation of the data and the accuracy of the regression model were always worse than that after excluding the soil background when the maize at the filling stage. The absolute value of the correlation coefficient between the reflectance of the soil background and the moisture content of each depth ranges from 0.01 to 0.33， and the absolute value of the correlation coefficient of the reflectance of the soil background and the moisture content of each depth ranges from 0.08 to 0.54. The effect of the model with soil background is higher at any depth than that of the model without soil background， and the accuracy of the partial least-squares regression model is the best at each depth.

The aboveground part of surface vegetation is mainly the stem of vegetation. The rhizome of vegetation can effectively reduce the sediment transport capacity and contribute to the water and fertilizer management of plants. In this study， U the rigid vegetation with different rhizome sizes was numerically studied by using Reynolds stress turbulence model and three-dimensional software FLUENT. Three kinds of vegetation with different rhizome sizes were simulated and designed， which were 3.5 mm， 5.0 mm and 6.5 mm respectively. At the same time， the same vegetation density was maintained and the experiment was carried out under the same flow conditions. The results show that in the submerged state， with the increase of vegetation coarseness， the velocity between vegetation decreases gradually， the velocity at the channel increases gradually， and the tail vortex and vortex increase gradually. The continuous increase of vegetation coarseness leads to the gradual decrease of average velocity. Compared with 3.5 mm coarseness， the average velocity of 5.0 mm and 6.5 mm coarseness decreases by 6.3% and 10.6% respectively. Turbulence kinetic energy and intensity are positively correlated with vegetation coarseness. These results are of great significance to optimize the distribution of slope vegetation in farmland protection.

Remote sensing quantitative monitoring of soil moisture in a large area is of great significance to contemporary precision agriculture applications. But how to improve the monitoring accuracy has always been a key issue in this field. The fusion method of multi-source remote sensing can give full play to the advantages of various remote sensing and is an important technology to improve monitoring accuracy. Taking the central and eastern part of Henan Province as the research area， based on MODIS and Sentinel data and measured soil moisture， according to the contribution of vegetation cover， surface roughness and soil moisture to backscattering， the BP neural network model was used to construct the relationship between the above parameters. And the soil moisture content in the depth of 0~10 cm and 0~20 cm during the high vegetation cover period of winter wheat in the study area from March to June， 2016， was respectively inverted. Depending on the properties of the surface roughness parameters， the hypothesis of constant surface roughness was put forward and combined with genetic algorithm (GA-BP) to optimize the BP neural network method to carry out comparison experiments. The results show that： ①In the period of luxuriant vegetation， the backscattering coefficient (σ) and its difference value (?σ) have a certain correlation with soil moisture. VV polarization is better than VH polarization， and the difference value is better than the original value. ②When inverting soil moisture in depths of 0~10cm and 0~20 cm， the accuracy of the results obtained by the BP _σ， BP_{? σ}， and GA-BP_{? σ} models are all improved sequentially， and the root mean square error of the GA-BP_{? σ} model is 4.07% at 0~10 cm and 3.42% at 0~20 cm. ③The three BP neural network models all have a better correlation with the soil moisture in the depth of 0~20 cm， and the prediction accuracy is high. The research shows that the assumption of constant surface roughness during the whole growth period of winter wheat in the Central Plains is valid， the backscatter coefficient difference (?σ) has a better correlation with soil moisture， and the remote sensing of root moisture content of 0~20 cm is more sensitive.

In order to study the effect of three-bands spectral index inversion of soil moisture content to provide a reference for the rapid and accurate determination of surface soil moisture content in precision agriculture， in this study， 159 soil samples of different content in Yongkang area of Zhejiang were collected and the spectral reflectivity of the collected clay was measured by utilizing ASD FieldSpec 3 spectrograph indoors. Based on the original spectral reflectance (R) and its corresponding first-order differential spectrum (FD) and second-order differential spectrum (SD)， one-band， two-band and three-band spectral index were extracted. Then， different spectral indices were modelled through partial least squares regression (PLSR) and the accuracy of soil moisture content predicted by the models was analyzed. The results indicate that： the three-band spectral index is more sensitive to soil moisture content than the one-band and two-band spectral indices； the partial least squares regression model based on the three-band spectral index (TBI ₂) has the best prediction effect， with R_v ²=0.92 and RPD up to 3.32； for soil moisture content inversion， the three-band index (R-TBI ₁) is more important than other spectral indices， with VIP value of 1.04. This study shows that the use of three-dimensional spectral index modeling provides a new way for hyperspectral remote sensing to quickly and effectively monitor surface soil moisture content.