Agricultural Meteorology
M. Abdollahi Fuzi; B. Bakhtiari; K. Qaderi
Abstract
IntroductionSpring frost is considered an important threat to agricultural products in high and middle latitudes. The damage caused by Late Spring Frosts (LSFs) significantly impacts vulnerable plant organs. This event has caused more economic losses to agriculture than any other climatic hazard in Asia, ...
Read More
IntroductionSpring frost is considered an important threat to agricultural products in high and middle latitudes. The damage caused by Late Spring Frosts (LSFs) significantly impacts vulnerable plant organs. This event has caused more economic losses to agriculture than any other climatic hazard in Asia, North America, and Europe. Also, these phenomena have contributed to low crop yields in Iran. The latest statistics released by the Food and Agriculture Organization of the United Nations (FAO) show that Iran is one of the largest producers of agricultural products and the world’s second-biggest producer of pistachios. Kerman province is one of the significant areas of pistachio production. This province has a large share of the pistachio word area plantation. Spring frost damage to pistachio crops has led to low yields in recent years. A key aspect of studying frost is the ability to accurately estimate its occurrence. In this study, artificial neural network methods have been used to estimate late spring frost in the pistachio crop of Kerman city. Materials and MethodsIn this study, the efficiency of this method was investigated in the estimation of minimum temperature. For this purpose, the daily data of the synoptic station of Kerman city were obtained from Iran Meteorological Organization from 2000 to 2020. Meteorological data including mean, maximum, and minimum temperatures, relative humidity, wind speed, saturated vapor pressure, and sunshine hours were used. Five different combinations of these variables was considered as input variables in artificial neural network method for minimum temperatures modeling. After entering data into network and modeling with each combination, RMSE and R2 values were calculated. Finally, the combination of 8 variables including average and maximum temperature, the minimum temperature the previous day and two days prior, relative humidity, wind speed, saturated vapor pressure, and sunny hours were selected as the most suitable combination of variables. Subsequently, a simulation of minimum temperature values was conducted using 10% of the data. The performance of the methods was evaluated using statistical indices of coefficient of determination (R2), mean square of error (RMSE), Mean Bias Error (MBE), and Coefficient of Nash–Sutcliffe (NSE). Results and DiscussionThe accuracy of an analytical method is the degree of agreement between the test results generated by the method and the true value. Upon examining the models, the M1 model was identified as the best due to its lowest RMSE and higher R². ANN model results were evaluated using various performance measure indicators. The simulated outcome of the model indicated a strong association with actual data, where the correlation coefficient was above 0.95, and the MBE index was zero. Also, the RMSE value was positive and close to zero, and the NSE value was above 0.75. Therefore artificial neural network method had high accuracy. In this study, mean annual minimum temperature was estimated using artificial neural network models (from March 10 to May 20). Comparison between the observed and calculated data showed that these data were in good agreement. Also, the results showed that temperature fluctuations were high between March 10 and March 31. From 2011 to 2017, an almost uniform temperature trend has been observed between March 10 and March 31. However, the years 2000, 2006, and 2020 showed a noticeable decrease in temperature. From 2018 to 2020, this trend of temperature reduction continued. In April, the temperature values were between 7 and 10 degrees Celsius. The years 2001, 2005, 2006, 2009, 2016, and 2019 had a noticeable decrease in temperature. In May, the mean minimum temperature was between 10 and 14 degrees Celsius. Therefore, the probability of frost occurrence in early-flowering cultivars was higher in late March than in April and May. The years 2000, 2004, 2005, 2012, 2015, 2019 and 2020 had the highest number of frost days in the last two decades. ConclusionThe results showed that the artificial neural network method had a high performance in estimating the minimum temperature. The values of the statistical indicators were R2=0.963, RMSE=0.027oC, MBE= 0 and NSE=0.966 respectively. In addition, the ANN method performed well in estimating the number of critical frost days for pistachio crops. The results showed that, although reducing the amount of input data in models decreases their output precision, data-driven methods can still be useful tools for minimum temperature estimation.
Agricultural Meteorology
S.M. Ebnehejazi; H. Yazdanpanah; S. Movahedi; M.A. Nasr-Esfahani; M. Moradizadeh
Abstract
Introduction Agricultural products frost in spring imposes heavy financial losses to agriculture particularly in northwest of Iran’s orchards. Not only temperature is one of the most important climate parameters but also it is a very crucial element in the agricultural sector. Untimely temperature ...
Read More
Introduction Agricultural products frost in spring imposes heavy financial losses to agriculture particularly in northwest of Iran’s orchards. Not only temperature is one of the most important climate parameters but also it is a very crucial element in the agricultural sector. Untimely temperature fluctuations and rise and fall which are usually unexpected will cause shock and heavy damages. Therefore taking into consideration the agricultural products frost and offering an approach would be of great importance for reducing relevant damages. In studies carried out by Omidvar and Dehghan Banadoki (2012) and Hesari et al. (2015) characteristics and different types of frosts have been considered in relation to the agricultural products. Different models were introduced to predict flowering date in different investigated regions. In more studies, in addition to determining the best model for predicting the date of occurance of flowering stage, probable date of last frost has been estimated as well. Investigating long term temperature changes is a method which applied by Martínez-Lüscher (2017) and Vitasse et al. (2018) to find out about established changes in flowering date and also changes in the last frost date. Nasr Esfahani and Yazdanpanah (2019) realized that 48-hour early warning for frost occurrence can be performed with adequate precision. Despite all studies in the field of products frost particularly during flowering date, it seems a rapid frost warning system must be established and provided to make early warning for each orchard. In this essay, since our goal is to make such early warning three days before frosting, so we have to investigate accuracy and validity of 72-hour minimum temperature simulation using WRF model. On the other hand, we must know phonological stage of each product in each orchard to inform the farmer about frost hazards based on critical temperature, therefore the second goal of this research is to detect phonological stages through Landsat 7 and Landsat 8 images.Materials and MethodsIn order to achieve the aim of current study, 72-hour minimum temperature simulation through the Weather Research and Forecasting (WRF) model was investigated and values of vegetation index were derived for a 30 meters pixel at an experimental orchard in Kahriz, West Azerbaijan Province, in 2016-2107. Computational grid for 2 meters temperature simulation using WRF model contains of three nested grid with horizontal resolution of 27, 9 and 3 kilometers. Horizontal resolution of terrain height and land use data is equal to 30 second (about 1 km). The initial and 3-h boundary conditions with 0.5º horizontal resolution from the Global Forecast System (GFS) were obtained from National Centers for Environmental Information (NCEI). Based on the previous research KFMYJ physical scheme configuration for WRF model were used in this research. Model's hindcasts at 03:00 UTC hour for each of 51 synoptic weather stations of northwest of Iran in internal computational grid were interpolated by MATLAB software with interp 3 function using linear method, then the obtained values were compared to minimum temperature observed in the stations by using MAE, MSE, RMSE and MSSS indicators. Phenological statistics, the time of beginning and end of growth stages were obtained from Iran Meteorological Organization. Besides, 77 Landsat 7 satellite images of ETM+ sensor, and 41 Landsat 8 images of OLI sensor were downloaded from United States Geological Survey website from March to September 2007-2016 with a spatial resolution of 30 meters. In this research, atmospheric and radiometric correction were performed with the FLAASH method on the metadata file in the ENVI software environment and then vegetation index was calculated using NDVI index.Results and DiscussionExamining the evaluation indicators of the WRF model, results revealed a significant correlation and regression model between 2 meters temperature variable from WRF model output and minimum temperature variable observed in the entire stations for 72-hour simulation. As a result WRF model can be applied in 72-hour temperature simulation in the area of study. Another finding of this research indicated that in comparison to the field-recorded data, NDVI values gained from Landsat images properly indicates changes of phenology stages in the relevant apple orchard. In this study, the indicators used to evaluate the model error showed model hindcasts are more accurate for 24-hour and then 48-hour simulations than for 72-hour simulation, but the 72-hour simulation accuracy is not much different from 24-hour and 48-hour simulations. In northwestern Iran, which is a mountainous region, it is very difficult to simulate airflow in areas with complex topography, therefore the total correlation coefficient of all stations in all three simulations is in the range of 0.5, and the error rates of MAE and RMSE, respectively reaches about 2.8 and 3.8 Celsius. According to the second finding of this research, the NDVI indicator obtained from Landsat 7 and Landsat 8 satellite images can show the progress and changes in the phenological stages of apple trees.Conclusion This study showed the efficiency of the WRF model for 72-hour simulation of the minimum temperature as well as the potential of Landsat 7 and Landsat 8 images in detecting apple phenological stages in the study area. Therefore, by using the WRF model for 72-hour minimum temperature simulation and recognizing the phenological stages from Landsat images, if the temperature in any orchard reaches a critical level in the next 72 hours due to the phenological stage, frost warning can be announced and then frost mitigation should be done by the farmer.
