The Simulation Of Temperature Space In The Cool-climbing Area And The Response Of Tobacco Leaf Quality To Temperature

Natural condition and human activities are controlling factors in air temperaturevariation. Meanwhilt, air temperatures have different values in different regions and points which affect plant growth. As study areas, Liangshan and Panzhihua are essential tobacco producing areas in China. The special climate and complex topography made it difficult to get the air temperature of each point in the region from a few meteorological stations. In order to solve this problem, the remote sensing products were used to establish inversion model. In this paper, simulation data of air temperatureswith high accuracywere obtained. Sixquality indexsof tobacco leaves (total sugar, reducing sugar, total nicotine, chlorine, potassium, total nitrogen) were measured in addition to analysis the correlation between air temperatures and the quality of tobaccos. Furthermore, the critical time point which air temperature had the maximam effect on corresponding index was found to provide theoretical basis for the ecological cultivation of flue-cured tobaccos in study areas. The main results are as follows: (1) statistical analysis of temperature actual measurement dataThe all regional temperature change trend is complex with annual temperaturevarybetween-6.99? and29.4?, stretch ing across 36.3?. The difference between the highest and the lowest temperature of is about 20?. The annual maximum temperature for the West is in late May, the middle part in late April, the East in early August, southern Panzhihua in late April, mid April, the lowest temperature is early February. The highesttemperature of all area isPanzhihua, then the annual averagetemperature from high to low are the temperature in Southern,middle, Eastern and Western parts.(1) Simulation and verification of spatial change of air temperatureFirstly, the air temperature (AT)data of meteorological station in 2014 were collected and the AT data set was set up. Secondly, corresponding land surface temperature (LST) data were retrieved by MODIS products and the LST data set was set up. Then, the simulation model of air temperatures of Liangshan and Panzhihua was established based on both two data sets. Finally, annual spatial simulation maps of every ten days came up with the 250×250 kilometers grids were charted and the verificationof the results are as follows:The annual mean absolute error(MAE) is 1.75?and the root mean square error(RMSE) is 2.15?. The mean absolute error(MAE) in early April to early September is 1.31?, and the root mean square error(RMSE) is 1.52?. The high simulation precisionmeets the research requirements. Panzhihua, the south and central part performedhigher temperatures, the west and east pat performed relatively lower temperatures.The temperature droppedfrom early January to late February andrised from early March to early in April.The area suffering a high temperature periodfrom mid April to late September andthe temperature dropped fromearly October to late December.(2) Statistical analysis of quality indexsof tobaccoIn the upper part of the leaf, the potassium content, potassium chloride ratio, total nicotine and total nitrogen were significantly different in each region, however, other indicators did not show significant differences.There were significant differences in the total nicotine, the ratio of sugar to nicotine, the ratio of potassium to chlorine, the ratio of potassium to chlorine and the ratio of onitrogen to nicotine in the middle leaves.Moreover, the contentof reducing sugar, total nicotine, the ratio of sugar to nicotine, the content ofpotassium, total nitrogen and the ratio of onitrogen to nicotinein the lower parts of leaves were significantly different.(4) Response of tobacco leaves quality to air temperatureThe correlationanalysis showed that most of indicators are significantly correlated with temperature.Total nicotine in the upper and middle part of the leaf shows significant negative correlation withthe temperaturein mid May. At the same time, the content of potassium inthe lower part of leaf showssignificant negative correlationwith the temperatureinmid May.Analysis of each region:Thecorrelationanalysis of the study areas showed that the indicators in the upper, middle and lower parts of tobacco leaves are significantly correlated with air temperature. In the west part, reducing sugar showed a significant negative correlation with air temperature in the upper leaf during April and mid June.During early June ofPanzhihua,chlorineshowed a very significant negative correlation and a very significant positive correlation with the air temperaturein the middle and lower part of leavesrespectively.Anylsis of all varieties:The content of total sugar, reducing sugarand nicotine in the upper leaves of Yunyan 87were significantly correlatedwith the temperaturein early August. The content of nicotine in the middle part of leaves was significantly correlated withthe temperature in early May, midJune, lateJune and late July.The contentof total nitrogen was significantly related withthe temperaturein early April, early May, mid June and last June.The indicators in each part of tobacco leaves were most affected by the temperature inAugust, for examplethe content of total sugar, reducing sugar, nicotine and total nitrogen in the upperpart of leaveswere significantly relatedwith the temperaturein early August.Besides, Total nicotine, chlorine, total nitrogen in the middle part of leavesweresignificantly related withthe temperature in mid August. The content of potassium and total nitrogen inthe lower part of leafwere significantly correlated with the temperature in late August.The content of total sugar, reducing sugar, nicotine, potassium, total nitrogenin the upperpart ofYunyan 85 leaves were significantly correlated with the temperature inmid May. The content of potassium in the middle part of Yunyan 85 leaveswas significantly correlated with air temperature in midJuly and early August.