| The research mainly focuses on irrigation prescription map construction of crops irrigated with largescale lateral moving sprinkler irrigation system.Firstly,the hydraulic performance of large-scale sprinkler irrigation machine and sprinkler operation parameters were tested.Then the spatial distribution of crop information,such as canopy temperature,crop water deficit index,plant height,leaf area and yield,was obtained using UAV spectral sensing system with data collected from thermal infrared,multispectral,visible light and other spectral images.Lastly,the irrigation prescription map was built with the inversion soil moisture data under the situation of variable sprinkler irrigation,combined with the typical annual irrigation information that retrieved from historical meteorological data.The constructed irrigation prescription map provides information that supports for precision and smart irrigation.Specifically,1.The running parameters of large-scale lateral moving sprinkler machine and the hydraulic performance of sprinkler nozzles were tested and simulated by using the method of spatial interpolation.The spraying effect of rotary sprinkler and the superposition effect of different sprinklers were simulated,and the relationship between the rate and the percentage was found to be:y=1.58x-3.1089,x?[1 0,1 00],R2=0.984,running speed was 32.2?158.6 m/h.The test results showed that the spray water quantity of different nozzles has the same variation trend under different running speed.With the increase of the running speed of the lateral sprinkler machine,the spraying water quantity of sprinkler nozzle decreased in the power function of y=a*x^ B.As the spray range and droplet size of different sprinkler nozzles are different,the nozzles should be selected according to different climatic conditions,physiological growth characteristics of irrigated crops and the demand of variable and precision irrigation.2.Irrigation management of winter wheat based on water deficit inversion with UAV remote sensing was benefit to improve the efficiency of intensive farmland management and the information acquisition efficiency of the water and fertilizer spatial distribution to achieve precision irrigation.In this research,three irrigation levels were set using different irrigation treatments with large-scale lateral moving sprinkler system.The spectral data of different test plots were obtained using UAV carried thermal infrared,visible light and multispectral cameras.Then the vegetation index was calculated and water and fertilizer indexes were retrieved.Moreover,a soil moisture inversion model was constructed combined with field sampling and calibration.Under the condition of water deficit,the amount of irrigation directly affects the physiological growth index of winter wheat,the more irrigation,the better development of plant height and leaf area.The difference of canopy temperature among the three irrigation levels was 2?5? and the vegetation index is consistent with the irrigation treatment.The crop water deficit index calculated by thermal infrared image inversion of crop canopy temperature can indicate the spatial distribution of water deficit of winter wheat.There is a good correlation between crop water deficit index and average soil moisture.3.In this study,the effect of measuring plant height and leaf area by UAV multispectral remote sensing in the situation of no point calibration was tested by using variable irrigated winter wheat with lateral moving sprinkler irrigation.Three irrigation levels of 240 mm(IT1),190 mm(IT2)and 145 mm(IT3)were carried out,and 60 sampling plots were arranged under each irrigation treatment.The plant height was measured by ground manual measurement and calculated with point-based data processing from image collected with UAV multi-spectral remote sensing.The results showed that 1)Under the three conditions of different irrigation treatments,the trend of plant height measured manually on the ground was consistent with that calculated by point-cloud processing of UAV multispectral remote sensing image,and the effect of irrigation water on plant height was positively correlated.The average plant height of the three irrigation treatments was IT1>IT2>IT3,and the daily growth rate after the jointing stage was IT1>IT2>IT3.2)The variation trend was consistent between the ground measured data and the multispectral point-cloud data,and the difference between the plant height extracted from point-cloud data was 12 cm for first three times and 15 cm in the later period.The plant height extracted from the point-cloud processing can effectively retrieve the differences of plant height and leaf area among different plots.Moreover,the plant height,leaf area index and other information can effectively extract from the point-could processing while collecting the multispectral images.4.Ten kinds of machine learning algorithms were used to estimate the yield of each plot using the spectral index calculated from multispectral and thermal infrared image data under three irrigation levels.The results showed that the average yield of each plot was 8.0 kg,6.7 kg and 4.9 kg,and the 1000-grain weight was 41.5 g,39.8 g and 35.5 g,respectively.The difference of yield gradient under the three treatments was more obvious than that of the 1000-grain weight,which indicated that the more irrigation amount was,the more wheat grains were.Based on the analysis of thermal infrared and multispectral data,the vegetation index yield prediction model was established.It was found that there were differences in the vegetation index under three irrigation treatments,and the significant differences were concentrated in T6,T7 and T8 periods.The Enhanced Vegetation Index(EVI)prediction model showed the most significant difference among the vegetation index prediction models.The correlation between vegetation index and yield was not significant in T1,T2,T3 and T4,but significantly increased in T5,T6,T7 and T8 periods.Compared with the yield prediction model made by thermal infrared data and without thermal infrared data,it is found that the overall accuracy was higher of the model made by using thermal infrared data.5.In this research,canopy temperature and crop water deficit index extracted from UAV spectral images were used as input for QWaterModel,then the spatial distribution image data of canopy evapotranspiration were obtained.Considering the characteristics of sprinkler irrigation and the physiological parameters in different crop growth periods,the inversion model of irrigation prescription map was developed.The results showed that 1)the spatial distribution of ET calculated by QWaterModel can well reflect the spatial variation characteristics of physiological growth activities of crop canopy.2)under the situation of insufficient meteorological data,the spatial distribution of crop canopy evapotranspiration can be retrieved by QWaterModel from crop canopy temperature image data that obtained through UAV sensing system.3)in the different irrigation treatments experiment,the spatial variation of irrigation treatment can be reflected by the irrigation prescription map,and 4)the spatial variation of the irrigation prescription map retrieved from the spectrum image will change with each variable irrigation process in the prosperous period of winter wheat.In this paper,the principle model of inversion irrigation prescription map mainly considers the available water in the soil.The upper limit of irrigation amount is the field capacity in the effective waterabsorbing root zone.The irrigation cycle is determined by crop canopy ETmax.It is necessary to further consider the critical crop canopy ET to determine whether irrigation is needed or not.The inversion method in this paper can provide a reference for modern precision irrigation information perception. |
PDF Full Text Request