Research Crop water use estimation
An annual water use ( or actual evapotranspiration , ET ) map was created for the entire South Africa for 2014 / 15 , with measures available for every 250mx250m pixel across South Africa . Spatial datasets from operational satellites , spatially interpolated weather data and other static datasets were generated in a pioneering manner using the ETLook 6 model . The water use was determined as the sum of transpiration , intercepted evaporation and soil evaporation estimates . Daily input data was used to generate 12 monthly maps , which were aggregated to an annual map . The resultant map of South Africa and extracts for the Western Cape and Mpumalanga provinces are shown in Figure 1 . The maps represent the consumptive water use across South Africa for the period 1 August 2014 to 31 July 2015 , expressed in mm / year .
Figure 2 : Actual monthly water use ( mm ) of irrigated and rainfed sugarcane ( left ) and rainfed winter wheat and irrigated table grapes ( right ).
The annual water use map ( s ) show areas with higher vegetation water use or ET ( green ) or lower water use ( red-orange ). The variations in the water use across the country are associated with water availability ( whether through rainfall or irrigation ), the vegetation or crop type and the climatic conditions . The map ( s ) show clearly how the water use by crops and other types of vegetation vary across the country and understanding this spatial variation in water use provides insights into water requirements in various regions where specific crops may dominate .
The monthly and annual water use maps generated in this project ,
Country wide mapping of irrigated agriculture
( 1 )… according to homogenous rainfall areas representing the actual water use , contain a wealth of information that can be explored further . The examples in Figure 2 show the actual water use over 12 calendar months ( 2014 / 15 ), effectively displaying the seasonality in a crops ’ water use or its “ water use profile ”. This monthly information is useful when comparing the
( 3 )... using information on crop water use , ( 6 )… together with machine learning eLEAF Biophysical methods
-Evapotranspiration ( ET )
Classification and
-Transpiration Regression Tree
- Interception
( CART ) - Evaporation - Biomass
( 4 )... difference between ET and rainfall ,
Rainfall ( P ) - ET minus P
Figure 3 : Ranges in the annual water use for a number of selected crops including irrigated and rainfed maize , irrigated and rainfed sugarcane , irrigated sugarcane and irrigated apples .
water use by different crops over the season ( e . g . dryland wheat vs . irrigated table grapes ) or to investigate the differences between irrigated and rainfed scenarios for the same crop type ( e . g . sugarcane ).
The water use information can also be extracted for multiple fields to provide insight into a specific crops ’ water use and water use requirements . Examples of the annual water use by maize , sugarcane , apples and citrus are shown in Figure 3 . These examples clearly illustrate the variation in water use per crop ; variation resulting from differences in water availability , efficiency of
PROVISIONAL
©
NOT FOR REPRODUCTION
( 2 )… at a field level using field boundary information
( 7 )… to map irrigated and rainfed fields
( 5 )... and multi-temporal vegetation indices Landsat-8 NDVI - Seasonal Maximum - Seasonal Minimum
What do you think of our results ?
8
Figure 4 : Process followed to generate the irrigated agriculture map .
SABI | APRIL / MAY 2017
Figure 5 : Map of actively irrigated agricultural areas in South Africa , as well as the rainfed areas , for the period 1 August 2014 to 31 July 2015