A paradigm of GIS and remote sensing for crop water deficit assessment in near real time to improve irrigation distribution plan
Introduction
A large part of world’s growing population is dependent on irrigated agriculture and it consumes more than half (about 56 %) of the global fresh water (Willer et al., 2008). Climate change and other factors including deteriorating water quality, depleting groundwater resources, and growing industrial and domestics demands are limiting water availability for irrigation (Rosegrant et al., 2010). Globally, gap between supply and demand of Irrigation water is expected to become high by the end of 2050, majorly caused by the population growth (de Fraiture and Wichelns, 2010d). Therefore, irrigated agriculture will be affected by concerns and issues over scarcity of water (Hu et al., 2016). A large amount of irrigation water get wasted or not managed efficiently to meet crop water requirement for achieving potential crop yield (Pereira et al., 2012). Thus, irrigation management authorities must aim to increase the expertise of monitoring and quantifying crop water demand, consumption and deficit rather than calculating supplied irrigation water only. This can improve water productivity and efficiency as well (Qu et al., 2015). Pakistan is one of the countries facing severe shortage of irrigation water although the country uses 90 % of fresh water for agriculture purposes (Rahut et al., 2016). Major sources of fresh water in the country including Himalayan’s glaciers and aquifers are depleting at enormous rate, while global temperature rise and decrease in effective precipitation intensify the water scarcity situation (Qureshi et al., 2010). Annual water availability in the Indus Basin Irrigation System (IBIS) is already towards decline and has squeezed from 130 Million Acre Feet (MAF) to 120 MAF by comparing average inflows of (1976–2000) and (2001–2016) respectively (Qureshi, 2011a). Canal Irrigation System, which feeds about 80 % of the agriculture land in Pakistan, is facing a rising trend in water shortages and is projected about 32 % shortfall in annual supplies by 2025 (Qureshi, 2011b). Canal water is supplied to the agriculture land on 08 daily rotation basis by implying the principal of equitable distribution. Fix amount of canal water per unit of land is supplied to the farmers irrespective of the type of crop grown (Shah et al., 2016). Rotation plan is developed at start of each crop season i.e. Winter (Rabi) and Summer (Kharif). Normally, secondary canals off taking from the main canal are divided into 03 groups and water is released in these groups on first, second and third priority on 08 daily bases. There is no scientific method adopted by the irrigation department to monitor canals fulfilling crop water demand, and rotation plan is followed inflexibly (Shah et al., 2016). There is a strong need to introduce new strategies and mechanisms to improve water use efficiency and productivity to ensure optimum use of irrigation water (Shah et al., 2006).
Knowledge of crop water demand, consumption and deficit are interlinked with each other and associated with crop specific evapotranspiration (ET) characteristics (Allen et al., 2011). ET requirements are typically expressed in terms of crop coefficient (KC) values for different crops, and their growing stages under various climatic conditions (Allen et al., 2001). To calculate ET, particularly Potential Evapotranspiration (PET), there exist a wide range of different field methods and empirical equations (Abdelhadi et al., 2000). (Djaman et al., 2019) compared and evaluated 34 equations of measuring PET by taking most worldwide acceptable and FAO’s recommended Penman-Monteith (P-M) method as reference. They concluded P-M equation of PET (PETPEN) demonstrated good results even under limited missing parameters like solar radiation, relative humidity and wind speed. Hargreaves equation (Hargreaves and Samani, 1985) is also well known and most simple method to calculate PET (PETHAR), but it underestimates and overestimates PET under wind speed more than 3 m/s and high relative humidity respectively (Allen et al., 1998). All of these methods are not suitable for spatial analyses, mainly due to limitations in spatial availability of required variables (Ahmad et al., 2009). An accurate assessment of ET and PET requires spatial detail of several parameters such as crop classification, climatic parameters, and Kc values at different growing stages and climatic conditions (Bastiaanssen et al., 2005). Therefore, RS datasets and GIS techniques are increasingly used to model various crop water requirement parameters in spatial domain and their monitoring in near real time manner due to wide coverage range and regular temporal data availability (Allen et al., 2007; Yousaf et al., 2018).
There are two basic approaches that are commonly proposed for monitoring water efficiency of irrigated crops thorough remote sensing (RS) data (Courault et al., 2005). First approach is based on Normalized Difference Vegetation Index (NDVI), in which reflectance-based crop coefficients (KCR) are obtained through regression calculations between local NDVI crop cycles and FAO defined crop coefficients (Glenn et al., 2011). Second method based on a series of satellite derived physical measurements, calculates actual evapotranspiration via surface heat budget (e.g. METRIC, ALEXI/DisALEXI, SEBAL) (Allen et al., 2005; Anderson et al., 2011; Bastiaanssen et al., 2005). Both methods have been successfully implemented in different regions of the world to estimate ET (Allen et al., 2007; Tasumi et al., 2005; Zwart and Bastiaanssen, 2007). A major requirement of the second method is ground station based meteorological data, whom availability and accessibility is usually limited particularly in developing countries (Tang et al., 2009). Although the estimation of ET from these schemes is very useful for planning studies where backdated assessments are required, but have relatively less potential in data limited regions for near real time decision making. (Peschechera et al., 2019) highlighted the adaptive capability of KCR in different climatic and specific field conditions as compared to standard crop coefficients approach. Crop water need or deficit (additional requirement) can be calculated as the difference of NDVI based ET of current crop (ETa) and the crop under ideal condition (PET) (Kumar and Rajan, 2014). Successful applications at basin scale have been reported from the Guadalquivir river basin in southern Spain (González-Dugo et al., 2013) and from Northern Italy (Rossi et al., 2010).
This research briefly highlights the use of RS and GIS technology to estimate crop water deficit in canal command areas (CCAs) at intervals parallel to the PID’s 8-day rotation plan. Water deficit monitoring on near real time basis can be helpful for irrigation managers to ensure better water allocation by making adjustments in ongoing rotation plan. Limitation of 08–10 days’ lag time in availability of satellite imagery will not impact much for making adjustments in canal water allocation while monitoring the cropping season spanned over 160–180 days. The main objective of this piece of work is to introduce flexibility in static rotation plans being practiced by PID since decades, by monitoring crop health using satellite technology based on open source data. The results can also help to promote guided groundwater abstraction by farmers incase canal water is insufficient to meet crop water demand, which would eventually reduce excessive groundwater pumping being considered as one of the major issue in the study area (Basharat and Tariq, 2014).
Section snippets
Study area
The study area chosen for this research is Bari Doab of Punjab Province in Pakistan. It is situated in IBIS and bears canal water shortages up-to 32 % (Nawaz et al., 2016). Bari Doab in Pakistan side of border exists between Ravi and Sutlej Rivers (as doab meaning land between two Rivers). It is one of the most fertile agriculture region of South Asia and is second largest irrigation system in Punjab province (Basharat and Tariq, 2014). The study area is intensely cultivated covering an area
Materials & methods
All the calculations regarding crop water deficit assessment were performed in Model Builder (Arc GIS 10.5®) using different spatial and statistical tools. Fig. 2 shows the research flow diagram highlighting used input datasets and various processing procedures adopted to calculate crop water deficit.
Crop water deficit assessment
The output of the model comes in the form of vector map, excel file and raster images for the whole investigation year on 8-day interval (46 investigations in total). Specimen of the PET (mm) and crop water deficit (mm) derived from the model in the form of vector map can be seen in Fig. 10. These results are average deficit values for each canal command in a particular 08-day period. Crop water deficit values at start of January are ranging from 0.05 to 0.30 mm for eight days, which is quite
Conclusions
In this research, it is concluded that Remote Sensing and GIS can provide repetitive, reliable and freely available data sources and methodology to assess and monitor crop water deficit on near real time basis. The research approach can be helpful in improving canals’ rotation plan by adjusting and allocating water from less demand towards more demand area. The methodology is based entirely on satellite-based imageries (i.e. MODIS, Landsat 8) and meteorological records, and both datasets are
Recommendations
It is recommended that future work may be focused on the forecast of crop water demand by incorporating climatic forecast one week ahead. This information in terms of predicted crop water demand be shared with all stake holders (i.e. agriculture department and irrigation department) to ensure better irrigation planning and scheduling. The knowledge of potential available sources of water (i.e. canal water, groundwater and precipitation) and crop water requirement in a particular duration would
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgments
The research is conducted under the project of Water Resource Management Information System (WRMIS) of Punjab Irrigation Department. The authors would like to thank Mr. Habib Ullah Bodla, Chief Monitoring of Decision Support System (DSS) and Program Monitoring and Implementation unit (PMIU), for initiating the project and encouragement. The authors would also like to express their gratitude to Mr. Roland Geerken from AHT, Germany, Mr. Khalid Mahmood and Mr. Tahir Mahmood Butt from NESPAK for
References (72)
- et al.
Estimation of crop water requirements in arid region using Penman-Monteith equation with derived crop coefficients: a case study on Acala cotton in Sudan Gezira irrigated scheme
Agric. Water Manage.
(2000) - et al.
Diagnosing irrigation performance and water productivity through satellite remote sensing and secondary data in a large irrigation system of Pakistan
Agric. Water Manage.
(2009) - et al.
Water productivity and crop yield: a simplified remote sensing driven operational approach
Agric. For. Meteorol.
(2018) - et al.
Satisfying future water demands for agriculture
Agric. Water Manage.
(2010) - et al.
Classifying rangeland vegetation type and coverage using a Fourier component based similarity measure
Remote Sens. Environ.
(2006) - et al.
Monitoring evapotranspiration of irrigated crops using crop coefficients derived from time series of satellite images. II. Application on basin scale
Agric. Water Manage.
(2013) - et al.
Estimation of actual irrigation amount and its impact on groundwater depletion: A case study in the Hebei Plain, China
J. Hydrol.
(2016) - et al.
Assessment of daily MODIS snow cover products to monitor snow cover dynamics over the Moroccan Atlas mountain range
Remote Sens. Environ.
(2015) - et al.
Improvements to a MODIS global terrestrial evapotranspiration algorithm
Remote Sens. Environ.
(2011) - et al.
Improved indicators of water use performance and productivity for sustainable water conservation and saving
Agric. Water Manage.
(2012)