M1 – Satellite remote sensing

Partner: UW in cooperation with UAF

a) Cotton
b) Rice
c) Wheat
d) Pre-Processing of MODIS NDVI data with bad quality pixels

Objective
Agriculture is the major water consumer in the irrigated Lower Chenab Canal of Punjab in Pakistan. The spatio-temporal inventory of major land use land cover classes is compulsory for successful assessment and optimal utilization of available water resources. Remote sensing data is key to generate such information, which is available free of cost nowadays from MODIS, Landsat and Sentinel etc., at a fine spatial and temporal resolutions.

Current Results
In the first instance, unsupervised land use land cover classification has been performed using MODIS NDVI data at 250 m spatial resolution. All major crops of rabi and kharif cropping seasons including wheat, rice, cotton, and sugarcane have been mapped for the period from 2005 to 2016. The accuracy assessment of the classified data is under process and finalized maps show overall encouraging results. The accuracy assessment is performed by using state owned crop inventory and ground-truthing data. Further actions of our research involve the estimation of crop specific consumptive water use and estimation of cotton yield by utilizing satellite data.

Preliminary Outcomes
All major crops of spring (Rabi) and monsoonal autumn (Kharif) cropping seasons including cotton, wheat, rice, and sugarcane are classified from 2005 to 2017. In this period, cotton was grown in Punjab province on 2.35 ± 0.21 million ha (47% of cultivated irrigated land). In LCC cotton is less dominant (15% of irrigated land), grown mainly in the tail end of the irrigation system. Accuracy assessment is performed by using multiple approaches including official crop inventory and own ground-truthing surveying 1,400 locations. Ongoing research estimates cotton yield, crop specific consumptive water use, and irrigation system efficiencies. Analysis and ground-truthing of a study site in Söke, Turkey is conducted for comparison.