Integration Remote Sensing and Meteorological Data to Monitoring Plant Phenology and Estimation Crop Coefficient and Evapotranspiration

The water requirements of the wheat crop are represented by the actual evapotranspiration, which depends on the meteorological data of the study area and the amount of water consumed during the season. Estimation of crop coefficients (Kc) and evapotranspiration (ETc) using remote sensing data is essential for decision-making regarding water management in irrigated areas in arid and semi-arid large-scale areas. This research aims to estimate the crop coefficient calculated from remote sensing data and the actual evapotranspiration values for the crop. The FAO Penman-Monteith equation has been used to estimate the reference evapotranspiration from meteorological data. Linear regression analysis was applied by developing prediction equations for the crop coefficient for different growth stages of comparing with the vegetation cover index (NDVI). The results showed that (R2 = 0.98) between field crop coefficient and crop coefficient predicted from (Kc = 2.0114 NDVI-0.147) in addition to (RMSE = 0.92 and (d = 0.97).


INTRODUCTION
Estimating the water requirement of agricultural crops is important for calculating and scheduling the amount of water needed for irrigation. Thus, it is the cornerstone in managing and preserving water resources. The most widespread methods for determining the actual water needs of agricultural crops depend on the use of the reference evapotranspiration and the crop coefficient curve (Abdalkadhum et al., 2020, Acharya andSharma, 2021). However, one of the most important disadvantages of this method is the difficulty of determining the yield coefficient curve and the lack of published data on this curve for all regions and for all agricultural crops. Remote sensing techniques have developed in the last decade were became used in various fields, the most important of which are agricultural applications, because they provide high accuracy in temporal and spatial information about the wide cultivated areas, types of agricultural crops, and it also provides a real monitoring system for the crop and its stages of development during the growing season for plants, through vegetation indices like (NDVI, SAVI, RVI, and DVI) ( Allen et al.,1998) have indicated the possibility of using these vegetation indices to calculate the crop coefficient. In central Arizona in the United States of America, a study was conducted (Hunsaker et al., 2003) in which the relationship between the crop coefficient for cotton and NDVI) retrieved from remote sensing data was analyzed. The relationship was initially linear with a correlation coefficient (R 2 = 0.97), later, it turned into a curve with a correlation coefficient (R 2 = 0.82) at the value (0.80) of the vegetative index (NDVI) which represents the end of the full coverage phase. The important challenges for agricultural horizontal extending policies and procedures in Iraq are limited water resources and water scarcity. At the same time, population growth is outpacing agricultural land availability. As a result, food quantity and quality diminish. To address this issue, policies of horizontal growth of agricultural fields and activities focused on saving irrigation water to agriculture in another region were implemented (Dingre, Gorantiwar and Kadam, 2021). Crop water requirements must be precisely assessed, irrigation efficiencies must be improved, and high-efficiency irrigation methods such as localized irrigation, scheduling irrigation, cultivating drought-tolerant crops, and cultivating shortterm varieties must be used. For water resource planning and irrigation management, data on crop evapotranspiration or consumptive water usage is critical (Rawat and Singh, 2016).
Managing scarce water resources to meet rising demands is trouble in such a scenario. The precise quantification of ETc at regional and local scales can assist in the development of water resource-based decision making and policy, as well as helping in the management of our water resources. ETc is an energy-driven process that is a significant component of the water budget  (Smith and Allen, 1996); the energy balance method (Allen et al., 2007; Aljanbi, Dibs and Alyasery, 2020); and the soil water balance method (Gibson, 2002) are some of the frequently utilized approaches. When extrapolating to a regional scale, the footprint of ETc measurement using the aforementioned methods is relatively smaller (Foken and Napo, 2008), which might cause significant bias. As for Iraq, no previous study was carried out on the use of remote sensing techniques to determine the yield coefficient curve. Therefore, the main objectives of the study are as follows: (1) Using the vegetation cover indicators obtained from Landsat 8 satellite images, estimating the crop coefficient (Kc) Yield evaporation (ETc) of wheat crop in the study area during different growing ages; (2) For different growth stages, mapping of crop coefficient and crop transpiration evapotranspiration across the entire field of study area.

Area of study
The research was carried out in the city of Al-Musayyib, which is located in the center of Iraq in the Babylon Governorate, between latitudes (32°30ˋ-32°50ˋ N) and longitudes (44°-44°20ˋ E) at a height of (32) meters above sea level. The total area of the city is 48936.701 hectares and the cultivated area is 25916.896 hectares, Figure 1. The city of Al-Musayyib is located on the eastern and western banks of the Euphrates River, which is divided into Al-Hindiya and Al-Hilla branches south of the city.

Climatic conditions
The average annual temperature in the city of Musayyib is (31℃), and temperatures reach their lowest values in January with an average of (-1 ℃) and reach their highest values in August with an average of (48 ℃), and the average annual rainfall is approximately (21 mm), and most of the precipitation occurs In the period between December and March, the relative humidity ranges between (36 %) in July to (88%) in February, and the wind speed in the city ranges between (1.6 m/s) in January and (3 m/s) in July, with an average of (2 m/s) (need reference).

Data collection
Meteorological data were collected from meteorological stations to calculate the reference evapotranspiration. Five of Landsat-8 satellite images (path 168/row 37) were downloaded from https://earthexplorer.usgs.gov/ around 10:30 a.m. local time with a resolution of 30 m. during winter and spring seasons were used to calculate normalized diff erence vegetation index NDVI. The satellite imageries were acquired on Jan. 3 th , 2020. Feb. 4 th , 2020. Mar. 32 th , 2020, Apr. 8 th , 2020, and May. 10 th , 2020.

METHODOLOGY
The relationship for both NDVI and Kc is observable. The Kc curve's similarity to a satellitederived vegetation index revealed the possibility of modeling Kc as a function of the vegetation index. As a result, the feasibility of determining Kc directly from a crop's satellite refl ectance was studied (Hubbard, 2013). Bands 4 and 5 of Landsat-8 provide red and near-infrared measurements, therefore they can be utilized to generate NDVI data using the formula (Rouse J., Schell J.A., Deering D.W., 1973): where: ρRed is the refl ectance in the 4 bands (red band); ρNIR is the refl ectance in the 5 bands (near-infrared band).
Crop coeffi cient values were determined based on the actual calculeted evapotranspiration (ETc) and the reference evapotranspiration (ETo) using the relationship: (2) The actual evapotranspiration (ETc) was estimated from the following water balance equation: The values of each of the surface runoff and water obtained by the plant by capillary action were considered equal to zero (the land is fl at and the groundwater is deep). The values of (ETo) were calculated using the equation (Penman-Monteith) (Allen, Pereira, Dirk Raes, et al., 1998; Acharya and Sharma, 2021) soil moisture, or groundwater. Over the years, various remote sensing-based surface energy balance (SEB based on the meteorological data at the In this study, shown in Figure (2), three stages were worked. The fi rst stage was the fi eld measurement of the crop coeffi cient and the calculation of (ETo) and (ETc) based on meteorological data and eqaution (2). In the second stage, fi nding a prediction equation between remote sensing data and crop coeffi cient data it was measured on the same day that the satellite images were acquired, and finally through the prediction equation and meteorological data (ETc) was calculated and compared with equation (3), then maps were made for each of the crop coeffi cient and (Kc) and actual evapotranspiration (ETc) on the periods of plant growth. Figure 3 shows the measured NDVI values of wheat during the growth stages. In general, the NDVI values were low in the early stages of plant growth, and the reason for the low value is that it is mainly related to the vegetative total of the plant (Gutman, 1991). Then the value increased with the growth stage of the plant and reaching the maximum vegetative growth during the fl owering stage in April. Then it decreased at the end of the season to a lower value, that is, in the stage of maturity. NDVI values ranged between (0.29-0.62).  the wheat plant in the study area were (0.29) for the plant stage during January, then increased to reach (0.44) in March in the vegetative growth stage and reached (0.63) in April during the maturity stage during May. We note that the values are less than (0.2) due to the presence of desert areas as well as the Hilla River within the study area. The results of the NDVI in Figure 3 shows, there is a variation in the density of vegetation cover during the study period, and the reason for this is due to the change in plant growth with time and the change in the prevailing climate in the study area, which has a signifi cant impact on plant growth, as the basic principle on which the NDVI indicator is based on strong refl ection of healthy plants based on chlorophyll at near-infrared wavelengths and its relatively weak refl ection in the visible red color. On the other hand shows the temporal variability in the average NDVI over the study period, as the red light is strongly absorbed by photosynthetic pigments (such as chlorophyll) present in green leaves, while near-infrared light passes or is refl ected through leaving the tissue, regardless of its color. This means that the areas with bare soil in the early stages of plant growth that have little or no green space cover are similar in refl ectance for both wavelengths. Figure 3 shows the NDVI maps for the period of the study area.  (2) and the crop coefficient was calculated from satellites. The lowest values of the crop coeffi cient were during the season in January, then this value started increasing until it reached the highest value during the fl owering stage (1.13) and then decreased during the maturity stage to reach (0.72) When comparing the crop values extracted from remote sensing data and the values of the actually measured crop coeffi cient and the values taken from the FAO data, it was found that the crop coeffi cient is (0.3, 0.4,0.42) for each of (KcFao, Kc-sat, Kccal.) respectively at the germination stage, then the values increase to reach (0.6, 0.5, 0.65) for each of (KcFao, Kc-sat, Kc-cal.) on successively at the stage of vegetative growth and continued to rise to reach values of more than 1, a stage in which the value of the actual evapotranspiration increases over the value of the reference evapotranspiration, as it reached (1.15, 1.06, 1.13) for each of (KcFao, Kc-sat, Kc-cal.) at the fl owering stage, but at the maturity stage and because of cutting irrigation at this stage, the values of the crop coeffi cient will return It decreased as it reached (0.4, 0.74, 0.72 ) at the maturity stage for each of (KcFao, Kc-sat, Kc-cal.) respectively. It is clear from the above data that there were no signifi cant diff erences between the values of (Kcsat, Kc-cal.) and they diff ered slightly from the values of (KcFao).

Crop coeffi cient
It is clear from the above that this factor varies according to the diff erent growth stages of the crop and climatic conditions, as it is affected by the factors that aff ect the plant density and the coverage ratio of the earth's surface, which are aff ected by the leaf area, the part of the land covered by vegetation cover, the age of the leaves, the opening and closing of the stomata, and the moisture of the soil surface, which were negatively aff ected by these characteristics. In irrigation, this was mentioned in the characteristics of vegetative growth that were aff ected under irrigation management. The FAO (1986) showed that the values of the wheat crop coeffi cient ranged from 0.3 to 1.87, depending on the climatic data of the agricultural season. The decrease in the values of the crop coeffi cient Kc at the stage of physiological maturity due to a decrease The crop's need for water due to the old age of the leaves and the low effi ciency of the roots in absorbing water, while the values of the crop coeffi cient Kc increased in the early stages of plant growth due to the concentration of water  (Fraenkel, 1986).

Study of the relationship between (Kc) and (NDVI)
The relationship between the crop coeffi cient Kc was fi eld-measured and (NDVI) calculated from satellite images in the same period was studied. Figure 5 shows a great agreement between the Kc and NDVI curves in terms of behavior during the growth stages, where the values of Kc and NDVI were low in the germination stage, then these values increased to achieve their highest values in the fl owering stage and then decreased to their lowest value before harvest.The correlation between the values of Kc and NDVI was also studied during the months from January to May, and the correlational relationship in the study area was (Kc = 2.0114NDVI -0.147 ) with a correlation coeffi cient (R 2 =0.96). This predictive equation is in agreement with (Er-Raki et al., Campos et al., 2010). The diff erence in the constants of the equation is due to the diff erent types of crop and climatic conditions in the study areas ( Figure 6). Figure 7 shows the daily crop evapotranspiration ETc values measured by the water budget method, the crop evapotranspiration measured by remote sensing data, and the values extracted from the FAO data. Where the data were low at the beginning of the season to reach the highest value (8.5 mm/day), while the maximum values  In can be seen from the results obtained, we can say that the actual crop evapotranspiration ETc of the crop varies according to the stages of growth, the number of days of the growth stage, the number of irrigations in it, and the increase in plant size, height and leaf area. The results show that the actual crop evapotranspiration values increased in the fl owering stage, due to the increase in plant height, increase in leaf area, and almost complete vegetative growth stage, which caused an increase in evapotranspiration (Figure 8).

CONCLUSIONS
The method of determining the crop coeffi cient (Kc) based on (NDVI) is a good way to determine the crop coeffi cient. It is a simple method compared to other methods used in determining (Kc) such as fi eld methods, or relying on the publications of the FAO, and one of its most important features is the ease of obtaining (NDVI) values, whether through satellite images such as Landsat-8 images or measurement using terrestrial sensors. But it should be noted that equation (2) is valid for the conditions of the study area and the studied crop only, and other studies must be conducted in diff erent areas on wheat or diff erent crops.