**3. Resources**

searchers such as D'Urso et al., (1992), Bastiaanssen (2000), Ambast et al., (2006) and Papada‐ vid et al., (2011) have highlighted the potentiality of multispectral satellite images for the appraisal of irrigation management. The integration of remotely sensed data with auxiliary ground truth data for obtaining better results is common in the literature. (Bastiaanssen et al., 2003; Ambast et al., 2006; Minaccapili et al., 2008). Ambast et al., (2006) have shown that the application of remote sensing data in irrigation is of high importance because it supports management of irrigation and is a powerful tool in the hands of policy makers. It has been found that research in ETc is directed towards energy balance algorithms that use remote sensing directly to calculate input parameters and, by combining empirical relationships to physical models, to estimate the energy budget components (Minaccapili et al., 2008; Papa‐ david et al., 2010; Papadavid et al., 2011). All the remote sensing models of this category are characterized by several approximations and need detailed experimental validations. Multi‐ spectral images are used to infer ETc, which is the main input for water balance methodsmodels. For estimations of ET, ground truth data (Leaf Area Index-LAI, crop height) and meteorological data (air temperature, wind speed, humidity) is needed to support this ap‐ proach. In nearly every application of water balance model, knowledge of spatial variations

The use of remote sensed data is very useful for the deployment of water strategies since it can offer a huge amount of information in short time, compared to conventional methods. Besides convenience and time reducing, remotely sensed data lessens the costs for data ac‐ quisition, especially when the area is extended (Thiruvengadachari et al., 1997). Although remote sensing based ETc models have been shown to have the potential to accurately esti‐ mate regional ETc, there are opportunities to further improve these models testing the equa‐ tions used to estimate LAI and crop height for their accuracy under current agro-

This Chapter discusses the implementation of the most widely used models for estimating ETc, the 'SEBAL' and 'Penman-Monteith' which are used with satellite data. Such models are employed and modified, with semi-emprical models regarding crop canopy factors, to estimate accurately ETc for specific crops in the Cyprus area under local conditions. Crop

The study area is located in the area of Mandria village, in the vicinity of Paphos Interna‐ tional Airport in Paphos District in Cyprus (Figure 1). The study area lying in the southwest of Cyprus is a coastal strip between Kouklia and Yeroskipou villages. The area is a coastal plain with seaward slope of about 2% and it consists of deep fertile soils made of old fine deposits. The area is dissected by three major rivers, the Ezousa, Xeropotamos and Diarizos. The area is almost at sea level (altitude 15 m) and is characterized by mild climate which provides the opportunity for early production of leafy and annual crops. The uniform and moderate temperatures, attributed to the permanent sea breeze of the area, and the relative

Water Requirements have been determined based on the evapotranspiration values.

in meteorological conditions is needed (Moran et al., 1997).

meteorological and soil conditions.

26 Remote Sensing of Environment: Integrated Approaches

**2. Study area**

#### **3.1. Field spectroradiometer**

The GER (Geophysical Environmental Research) 1500 field spectroradiometer (Figure 2,3) is a light-weight, high performance, single-beam field spectroradiometer. It is a field portable spectroradiometer covering the ultraviolet, visible and near-infrared wavelengths from 350 nm to 1050 nm. It uses a diffraction grating with a silicon diode array which has 512 discrete detectors and provides the capability to read 512 spectral bands.

The instrument is very rapidly scanning, acquiring spectra in milliseconds. The spectroradi‐ ometer provides the option for stand-alone operation (single beam hand-held operation) and the capability for computer assisted operation through its serial port, which offers near real-time spectrum display and hard disk data transfer. The maximum number of scans (512 readings), can be stored for subsequent analysis, using a personal computer and GER li‐ censed operating software. The Lens barrel used for the specific spectroradiometer is the Standard 4 field of view. The data are stored in ASCII format for transfer to other software.

inactive (Papadavid et al., 2011). Reflectance was calculated as a ratio of the target radiance to the reference radiance. The target radiance value is the measured value taken on the crops (Figure 3) and the reference radiance value is the measured value taken on the standard Spectralon panel (Figure 2), representing the sun radiance, which reaches the earth surface

Remote Sensing for Determining Evapotranspiration and Irrigation Demand for Annual Crops

http://dx.doi.org/10.5772/39305

29

**Figure 3.** Spectroradiometric measurements over potatoes (target) in Mandria Village in Paphos, Cyprus (Papada-

Leaf Area Index is commonly used for monitoring crop growth. Instead of the tradition‐ al, direct and labor-consuming method of physically measuring the plant with a ruler (direct method), an optical instrument, SunScan canopy analyser system (Delta-T Devices Ltd., UK) is used (indirect method). The instrument (Figure 4) is indirectly measuring LAI by measuring the ratio of transmitted radiation through canopy to incident radia‐ tion (Figure 5). Indirect methods for LAI measurements based on the transmittance of ra‐ diation through the vegetation have been developed (Lang et al., 1991; Welles and

—without atmospheric influence.

vid, 2012)

Norman, 1991).

**3.2. SunScan canopy analyser system**

**Figure 2.** Spectroradiometric measurement over spectralon panel (Papadavid, 2012)

Reflectance factors using a control stable surface with known characteristics as described by McCloy (1995) have been measured. Many researchers (McCloy, 1995; Beisl, 2001; Anderson and Milton, 2006; Schaepman, 2007; Papadavid 2012) highlight the advantages of using con‐ trol surfaces in the measurement of reflectance factors (Bruegge et al., 2001). In this study, the control surface was a commercially available "Labsphere" compressed "Spectralon" white panel (Figure 2). There is evidence that these types of panels are more consistent and retain their calibration better than painted panels (Jackson et al., 1992; Beisl, 2001). Spectra‐ lon diffuse reflectance targets are ideal for laboratory and field applications such as field val‐ idation experiments, performed to collect remote sensing data due to the fact are: durable and washable; have typical reflectance values of 95% to 99% and are spectrally flat over the UV-VIS-NIR spectrum; are impervious to harsh environmental conditions and chemically inactive (Papadavid et al., 2011). Reflectance was calculated as a ratio of the target radiance to the reference radiance. The target radiance value is the measured value taken on the crops (Figure 3) and the reference radiance value is the measured value taken on the standard Spectralon panel (Figure 2), representing the sun radiance, which reaches the earth surface —without atmospheric influence.

**Figure 3.** Spectroradiometric measurements over potatoes (target) in Mandria Village in Paphos, Cyprus (Papadavid, 2012)

#### **3.2. SunScan canopy analyser system**

and the capability for computer assisted operation through its serial port, which offers near real-time spectrum display and hard disk data transfer. The maximum number of scans (512 readings), can be stored for subsequent analysis, using a personal computer and GER li‐ censed operating software. The Lens barrel used for the specific spectroradiometer is the Standard 4 field of view. The data are stored in ASCII format for transfer to other software.

28 Remote Sensing of Environment: Integrated Approaches

**Figure 2.** Spectroradiometric measurement over spectralon panel (Papadavid, 2012)

Reflectance factors using a control stable surface with known characteristics as described by McCloy (1995) have been measured. Many researchers (McCloy, 1995; Beisl, 2001; Anderson and Milton, 2006; Schaepman, 2007; Papadavid 2012) highlight the advantages of using con‐ trol surfaces in the measurement of reflectance factors (Bruegge et al., 2001). In this study, the control surface was a commercially available "Labsphere" compressed "Spectralon" white panel (Figure 2). There is evidence that these types of panels are more consistent and retain their calibration better than painted panels (Jackson et al., 1992; Beisl, 2001). Spectra‐ lon diffuse reflectance targets are ideal for laboratory and field applications such as field val‐ idation experiments, performed to collect remote sensing data due to the fact are: durable and washable; have typical reflectance values of 95% to 99% and are spectrally flat over the UV-VIS-NIR spectrum; are impervious to harsh environmental conditions and chemically

Leaf Area Index is commonly used for monitoring crop growth. Instead of the tradition‐ al, direct and labor-consuming method of physically measuring the plant with a ruler (direct method), an optical instrument, SunScan canopy analyser system (Delta-T Devices Ltd., UK) is used (indirect method). The instrument (Figure 4) is indirectly measuring LAI by measuring the ratio of transmitted radiation through canopy to incident radia‐ tion (Figure 5). Indirect methods for LAI measurements based on the transmittance of ra‐ diation through the vegetation have been developed (Lang et al., 1991; Welles and Norman, 1991).

**3.3. Satellite imagery**

cycle monitoring.

**4. Methodology**

band, 15m for panchromatic and 30m for the rest).

Height (CH) to one of the selected Vegetation Indices (VI).

Spatial, spectral and temporal resolution of satellite images is very important for studies dealing with crop water requirements. Landsat- 5 TM and -7 ETM+ have been widely used for hydrological studies due to their relatively good temporal resolution (16 days) which is important for providing regular snapshots during the crop growth season (Dadhwal et al., 1996; Song et al., 2001; Alexandridis, 2003). These sensors are suitable for agricultural areas with medium to big average fields due to their high spatial resolution (60m for thermal

Remote Sensing for Determining Evapotranspiration and Irrigation Demand for Annual Crops

http://dx.doi.org/10.5772/39305

31

However, the availability of images is highly dependent on weather conditions. The availa‐ bility of cloud free images for operational projects is very important and depends on the geographical position and the prevailing weather conditions for the area of interest (Kontoes and Stakenborg, 1990; Hadjimitsis et al., 2000; Hadjimitsis et al., 2010). Countries such as Greece and Cyprus are characterised by good weather conditions and the availability of cloud-free images (Hadjimitsis et al., 2000). An advantage of Landsat image for applications in Cyprus is that of the whole island coverage from a single image which can be inferred on a regular basis since Landsat satellites overpass Cyprus on a systematic basis (Papadavid, 2011). Remote sensing users or policy makers or governmental officers have the oppotunity to have remotely sensed data very often which is very important in terms of phenological

D' Urso (1995) and Minacapilly et al., (2008) explored the importance of using image time series due to the high importance of water requirements in the different stages of the crops. The same crop in different stage has different water needs, therefore the time ser‐ ies of satellite images is very important in studies regarding ETc and remote sensing. A time series of Landsat 5 TM and 7 ETM+ imagery acquired in years 2008, 2009 and 2010 are used in this study, as listed in Table 1. The crucial aim is to have satellite images in all stages of the specific crops. The availability of images is important since these images will be converted into ETc maps using an image processing software such as ERDAS Imagine software. Hence, the more images we have the better analysis we get. All im‐ ages were pre-processed in order to remove atmospheric and radiometric effects, using the ERDAS Imagine software. ERDAS 'modeller module' was used to transform the im‐ ages into maps by applying the ETc algorithms. The same satellite images were also used for evaluating the statistical models found, regarding Leaf Area Index (LAI) or Crop

An attempt has been made to statistically describe the crop canopy factors, namely crop height (CH) and LAI, through the vegetation indices (VI). Crop canopy factors are vital ele‐ ments in the procedure of estimating ETc. These indices were produced from spectroradio‐ metric measurements using a hand-held field spectroradiometer (GER 1500) and after this data were filtered through the Relative Spectral Response (RSR) filters of the corresponding

**Figure 4.** SunScan (Delta-T) canopy analyser for LAI and crop height measurements

**Figure 5.** Use of SunScan canopy analyzer for LAI measurements (Papadavid, 2012)

#### **3.3. Satellite imagery**

**Figure 4.** SunScan (Delta-T) canopy analyser for LAI and crop height measurements

30 Remote Sensing of Environment: Integrated Approaches

**Figure 5.** Use of SunScan canopy analyzer for LAI measurements (Papadavid, 2012)

Spatial, spectral and temporal resolution of satellite images is very important for studies dealing with crop water requirements. Landsat- 5 TM and -7 ETM+ have been widely used for hydrological studies due to their relatively good temporal resolution (16 days) which is important for providing regular snapshots during the crop growth season (Dadhwal et al., 1996; Song et al., 2001; Alexandridis, 2003). These sensors are suitable for agricultural areas with medium to big average fields due to their high spatial resolution (60m for thermal band, 15m for panchromatic and 30m for the rest).

However, the availability of images is highly dependent on weather conditions. The availa‐ bility of cloud free images for operational projects is very important and depends on the geographical position and the prevailing weather conditions for the area of interest (Kontoes and Stakenborg, 1990; Hadjimitsis et al., 2000; Hadjimitsis et al., 2010). Countries such as Greece and Cyprus are characterised by good weather conditions and the availability of cloud-free images (Hadjimitsis et al., 2000). An advantage of Landsat image for applications in Cyprus is that of the whole island coverage from a single image which can be inferred on a regular basis since Landsat satellites overpass Cyprus on a systematic basis (Papadavid, 2011). Remote sensing users or policy makers or governmental officers have the oppotunity to have remotely sensed data very often which is very important in terms of phenological cycle monitoring.

D' Urso (1995) and Minacapilly et al., (2008) explored the importance of using image time series due to the high importance of water requirements in the different stages of the crops. The same crop in different stage has different water needs, therefore the time ser‐ ies of satellite images is very important in studies regarding ETc and remote sensing. A time series of Landsat 5 TM and 7 ETM+ imagery acquired in years 2008, 2009 and 2010 are used in this study, as listed in Table 1. The crucial aim is to have satellite images in all stages of the specific crops. The availability of images is important since these images will be converted into ETc maps using an image processing software such as ERDAS Imagine software. Hence, the more images we have the better analysis we get. All im‐ ages were pre-processed in order to remove atmospheric and radiometric effects, using the ERDAS Imagine software. ERDAS 'modeller module' was used to transform the im‐ ages into maps by applying the ETc algorithms. The same satellite images were also used for evaluating the statistical models found, regarding Leaf Area Index (LAI) or Crop Height (CH) to one of the selected Vegetation Indices (VI).
