Estimating evapotranspiration and irrigation water requirement in phonology stages of grape (VitisVinifera) using agro-hydrological model and remote sensing techniques

40 This study aims to compare the remote sensing (RS) approach and an agro-hydrological model to estimate 41 evapotranspiration (ET) and irrigation water requirement (IWR) in semi-arid region, and the effect of vineyards 42 management and their ages on these parameters. In the study region, after vineyards were classified into three 43 main scenarios based on three vineyards ages (12-15, 15-18 and 18-21 years) and two management approaches 44 (proper and improper management), ET and IWR were determined in each scenario using the Soil-Water- 45 Atmosphere – Plant (SWAP) Model and Surface Energy Balance Algorithm for Land (SEBAL) for the year of 46 2019-2020 with Landsat8 images. While the accumulated ET calculated with SEBAL was compared with a field 47 water balance, the results showed that without calibration or parameter optimization, the accumulated ET 48 estimated with SEBAL exceeded that computed with SWAP. According to the findings, the most and least 49 RMSE was related to August (1.32) and June (1.26). Analyses of scenarios showed that at the first stage of 50 phonology (bud-break to bloom), the S3 scenario has the most IWR for each pixel (900 m 2 ) by 2.7 m 3 , and at 51 the second stage (bloom to ripening) and the third stage (ripening), the S1 scenario by 229.5 m 3 and 78 m 3 has 52 the highest IWR, respectively.

between surface temperature and remotely sensed vegetation indices, WUE, evapotranspiration, and the 125 accuracy of remote sensing to estimating irrigation water requirement for phonology stages of VitisVinifera 126 with respect to SEBAL algorithm.

127
Actual evapotranspiration values obtained from the SWAP model was considered as the reference to be used for 128 remote sensing models accuracy assessment. Daily climate data (incoming short-wave solar radiation, air-129 temperature, humidity, wind speed, and rainfall) were obtained from Malayer synoptic station.

130
Field water capacity was determined in situ with a tensiometer, and the water content determined in the 131 laboratory was taken as permanent wilting point and the available soil water viz, then the quantity of water 132 remained in the soil between field capacity and permanent wilting point was computed. images were applied because of their temporal condition, high quality, lack of cloud cover, and easy access. The 141 selected images were level 2, which did not need geometric correction; in addition, they were checked by controller point (GPS) in the field, and atmospheric correction was done using FLAASH model.

156
The amount of net radiation flux is calculated using the input and output radiation waves according to equation Where RS ↓ is the short-wave input radiation (W / m 2 ), RL ↓ is the long-wave incoming radiation (W / m 2 ), RL ↑ is 159 the outgoing long wave (W / m 2 ), α surface albumin, and ε0 surface radiation strength of bands 10 and 11 of 160 Landsat8.   Where Ts is the surface temperature (˚C), and α is surface albumin. G value is obtained by multiplying the λET 172 in Rn. Normalized Difference Vegetation Index (NDVI) value less than zero is supposed to be water, and the G /

175
Sensible heat flux is amount of heat wastes through molecular conduction and convection, which is due to 176 temperature difference (eq.5) (Allen et al., 2002).

191
The numerical value of λ must be calculated through dividing the related number by per pixel to obtain the ET 192 value. By using instantaneous latent heat flux, the instantaneous ET value is obtained as follows (eq.6) (Allen et 193 al., 2002).

194
)6( = 3600 ETinst is the instantaneous ET (mm / hr) and λ is the latent heat ET (J / Kg), the value of λ is estimated by   234 235 Where kgr is the extinction coefficient for global solar radiation (Campbell and Norman, 1998

263
Water requirement is amount of water that is required for agriculture crops at the growth period, and it should be 264 supplied by irrigation. This amount is calculated by equation (16)   Due to the fact that surface cover has a great influence on the amount of ET in the surface of the region, and 314 SEBAL algorithm has a specific sensitivity to this parameter; therefore, the relationship between vegetation, 315 which is used to represent this component of the normalized differential vegetation index, and monthly ET was 316 evaluted. In this way, 500 random points were selected on the surface of two monthly ET maps, and vegetation 317 index on the day of satellite pass over the area and related correlations were calculated (Fig. 2). For each month, 318 the relationship between surface cover and the amount of ET was calculated to provide visibile results.  respectively. The field study showed that, in these months vines encounter with tension from a phenology stage 338 (bud-break to bloom) to another (bloom to ripening) so evapotranspiration would decrease.   In the LULC map of Malayer county, non cover and arable crop, vineyards (S1, S2, and S3), and the areas 363 covered by the specified orchards are cleared. According to the specific field studies, it has become clear that 364 the vineyard management strongly affects the type of vineyard, orchards, and the classification of the effective 365 factors. In this way, the legend of lifetime can be seen in Table 4. According to field study and Landsat images, 366 types of vineyards' management (in vineyards with the same age) have sensible differences in tone, texture, and 367 reflections from OLI sensor, so remote sensing techniques can be used to provide management map of 368 vineyards.    (Table 6). Effective rainfall of Non cover and 389 arable crops, vineyards (S1, S2, and S3) were 70.25 mm, and IWR of them were estimated less than 20 mm for 390 each class. However, the most IWR was observed by more than140 mm for the orchard gardens.  According to Table 7, statistical characteristics were estimated regarding IWR in GIS using IWR map and Table 7. Statistical characteristics of IWR from bud-break to bloom stage of vineyards 396 397

391
As shown in Table 7, IWR of Non cover and arable crops, vineyards (S1and S2) were zero; in fact, all their 398 IWR for growth were achieved through effective rainfall in 2020. Nevertheless, IWR of S3 was not provided 399 through effective rainfall (bud-break to bloom), and an average of irrigation water requirement is 2.83 mm, 400 meaning that IWR for per pixels of S3, was 2091.37 mm.  According to the analyses, it has become clear that more variation and dispersion of IWR of vineyards were 409 formed from bloom to ripening stage in comparison with bud-break to bloom stage. In this manner, pixels of 410 non cover and arable crops are as required higher IWR, which is related to seasonal crops based on field study, 411 than previous stage. As shown in Table 8, land uses are classified in five groups including non cover and arable 412 crops, vineyard (S1), vineyard (S2), vineyard (S3), and garden, and the highest mean weight IWR were related 413 to S1, S2, garden, S3, and non cover and arable crops areas, respectively.  Table 9 shows maximum, minimum, and standard deviation(STD) of IWR, total IWR, and IWR of per pixel for 417 different land uses from bloom to ripening stage. According this the biggest areas are belonged to S1, S2, non 418 cover and arable crops, S3, and garden, respectively, and non cover and crops and vineyard (S1) showed the 419 biggest and smallest STD by 0.158 and 0.12, individually. Besides, the highest and least total IWR were seen in 420 vineyard (S1) and garden, respectively, and the most IWR for per pixel were observed in vineyard (S1), 421 vineyard (S2), garden, vineyard (S3), and non cover and arable crops, respectively, and the maximum IWR was 422 founded in garden by 489.04 mm. According to Table 10, the most and least mean weight of IWR were founded in vineyard(S1) and non cover 432 and arable crops, individually, and non cover and arable crop, vineyard (S1), vinyeard (S2), vineyard(S3), and 433 garden showed the highest frequency of pixel in lower than 20 mm, 100-120 mm, 100-120 mm, 60-80 mm, and 434 80-100 mm, respectively.