In South Africa water is a critical resource for which there is strong competition between the urban, industrial and agriculture sectors. The National Water Act (NWA) of 1998 states fresh water should be used efficiently. The NWA prioritises water use for basic human needs and for protecting aquatic eco-systems with agriculture having a lesser priority. Nonetheless agriculture is of high economic importance as it contributes to food security, export, employment and livelihood.

Farmers in the Western Cape Province of South Africa have to cope with this increased competition for water for irrigation with other sectors while rainfall is also becoming less predictable due to climate change. Irrigated agriculture in the Western Cape is challenged to increase agricultural output (production) while reducing the amount of water used during the production, i.e. farmers need to increase their water use efficiency. With this background the Department of Agriculture: Western Cape funded the development of an open web portal where fruit and grape growers in the Western Cape have access to satellite based information on their crop growth and water use namely FruitLook.

FruitLook provides weekly updates on nine growth parameters such as actual evapotranspiration, biomass production and leaf nitrogen content that can help farmers understand the effects of farm management on their crop. On FruitLook, wine and fruit farmers can register (delineate) their irrigation blocks and analise crop growth and water status over time and space during the growth season. FruitLook data is provided at 20 m x 20 m spatial resolution that can show detailed variation across an irrigation block. In Figure 2 a screenshot of the FruitLook website is provided showing an overview of a table grape farm in the Hexriver Valley. Variations in biomass production between the different blocks and within the blocks themselves are visible.

FruitLook can help clarify why yields in one field are higher than in the other or what the effect is of irrigation scheduling on crop growth. Simultaneously FruitLook initiates conscious use of fresh water supplies in the Western Cape.

The FruitLook service has been operational since the 2011/12 season and is currently running for the third season in a row. The dataset behind FruitLook presents huge opportunities for statistical analyses of wine and fruit cultivation in the Western Cape. In this article a brief example evaluation based on FruitLook data is provided with a focus on table grape cultivation in the Hexriver Valley during the 2011/12 and 2012/13 seasons. Similar types of analyses could be applied to other areas and other crops within the FruitLook area as well. This includes statistical analyses of wine grape production, deciduous fruit production or other fruit farming sectors.


FIGURE 1. The average actual seasonal evapotranspiration (mm) per zone in the 2012/13 table grape season in the Hexriver Valley. A zone consists of an area with the same type of crop.

FIGURE 2. Hexriver Valley table grape blocks on FruitLook. Data is provided on 20 x 20 m spatial resolution enabling the user to view variation within and between different blocks. For this article the pixel data is simplified to field level.
FIGURE 3. Hexriver Valley seasonal actual evapotranspiration ET (mm) per field in the 2012/13 table grape season.
FIGURE 4. Hexriver Valley seasonal biomass WUE (kg/ha biomass per m3 water) per zone in the 2012/13 table grape season.
FIGURE 5. Histograms of actual evapotranspiration (actual ET) and biomass water use efficiency (biomass WUE) in the Hexriver area. The 2011/12 and 2012/13 seasons are depicted. The average actual ET and biomass WUE are showed above the light coloured column.
FIGURE 6. Biomass water use efficiency (kg/m3) per cultivar type in Hexriver Valley. Only cultivars grown on 50 fields or more were taken into account.


Using FruitLook data

The cultivation of table grapes requires a hot, dry climate with water supplied through irrigation. At the same time, water shortages can lead to direct reductions in table grape yield. Especially during certain developmental stages, table grapes are vulnerable for yield losses due to water deficits. Sufficient irrigation is critical for the commercial cultivation of table grapes and as a consequence the table grape industry is a large consumer of fresh water resources.

In the Hexriver Valley table grapes are the most common crop. Hexriver Valley is an ideal location for examining the possibilities of the FruitLook datasets in evaluating agricultural water consumption in larger areas. Seasonal actual evapotranspiration and seasonal biomass water use efficiency was extracted from the FruitLook dataset for the 2011/12 and 2012/13 seasons on a field by field basis. The pixel based data was simplified to field level for this analysis. The actual evapotranspiration (in mm) is the actual quantity of water lost from the land surface through evaporation of water from the soil and transpiration through plants. The biomass water use efficiency (WUE) signifies the total amount of biomass produced (in kg/ha) per cubic meter of water lost through evapotranspiration. A higher WUE value means the farmer is more efficient in using available water resources, implying efficient irrigation water use throughout the growth season, since rainfall is minimal in summer.

The examples provided in this article include: 1) determination of spatial patterns of water use in the Hexriver area, 2) comparison of data from multiple seasons, and 3) evaluation of crop varieties in terms of biomass water use efficiency.

1. Spatial distribution of water use

In Figure 1 the evapotranspiration data is congregated to zones. These zones represent areas with similar crop types. From Figure 1 it is clear that large variation exists within the Hexriver Valley. Figure 3 shows the distribution of water lost through evapotranspiration per field in the Hexriver Valley in 2012/13. In the western part of the valley, seasonal actual ET is higher than on the eastern part. This can be related to the convergence of the Hexriver and Amandel River in the western part. Furthermore, in the fields closer to the Hexriver, the actual evapotranspiration from the table grapes is mostly higher in comparison with fields more uphill. In 2011/12 (not depicted) the same distribution of water consumption is visible. From these observations it can be assumed that taking measures for lowering the water use in the western part of the valley will have more impact compared to the same measures taken for the eastern part of the valley.

In Figure 4 the distribution of seasonal biomass WUE is depicted per zone. The zones are similar to the zones depicted in Figure 1. The distribution of biomass WUE in the Hexriver Valley does not necessarily correspond to the distribution of actual evapotranspiration, i.e. high usage of water through evapotranspiration does not necessarily correspond to a low biomass WUE and vice versa. It does show there is a large variation within biomass WUE present within one crop type under similar climatic conditions, which can be illustrative for occurring over and under irrigation within an area. For example, a low biomass WUE in combination with high seasonal actual evapotranspiration is a strong indication for over irrigation. By combining crop evapotranspiration and biomass WUE a first evaluation can be made of efficient water use, per field or per zone, within the valley.

2. Seasonal variation in crop water use

The histogram distribution of actual evapotranspiration and average biomass water use efficiency in 2011/12 and 2012/13 per field in the Hexriver area is depicted in Figure 5. It is clear that during 2012/13 more water was lost through evapotranspiration by table grapes in Hexriver Valley. The biomass WUE is higher on average in 2011/12. The average in 2011/12 is 3.7 kg/ha/m3 biomass produced per water lost through evapotranspiration. In 2012/13 this was lower at 3.6 kg/ha/m3.

Higher seasonal accumulated evapotranspiration can be caused by climatic differences between the seasons. According to meteorological station data from the Hexriver area, precipitation amounts in the winter, before and during the beginning of the 2012/13 season, were higher than during the 2011/12 season. This could imply the water supplies stored during the winter and thus needed for irrigation during the summer months were substantially lower before the start of the 2012/13 season, and the farmers could irrigate consequently. Evapotranspiration in 2012/13 was further enhanced through higher air temperatures especially in the beginning of the growth season.

Through evaluation of multiple seasons of data, the biomass water use efficiency could be used to determine per block, whether the available water is used efficiently. A benchmark value could be determined to assess efficient irrigation water use by the table grape farmers. According to the biomass WUE histograms in Figure 5, it could for example be stated that a biomass WUE >4.0kg/ha for table grapes is good. This corresponds to the best ±30% of all table grape blocks in the Hexriver Valley in the 2011/12 and 2012/13 growth seasons.

3. Cultivar biomass water use efficiency evaluation

In Figure 6 the average seasonal biomass WUE (kg/m3) for various table grape cultivars in the Hexriver Valley is depicted. Only cultivars which are grown on more than 50 fields in the Hexriver Valley are taken into account. In this figure it can be seen that Crimson is (actually by far) the most biomass water efficient table grape according to the FruitLook data of both 2011/12 and 2012/13. Thompson is the most water efficient white table grape cultivar. According to the data Dauphine, Sultana and Victoria table grapes are less water efficient.

In literature Crimson is renowned for its vigour which corresponds to the results derived from the FruitLook dataset showing Crimson as the most vigorous table grape cultivar in the Hexriver Valley area. Crimson produces the most biomass per consumed unit of water. It must be understood that biomass does not equal yield. The biomass production encompasses dry matter growth of roots, shoots, fruits and all in between. Thus, while the Crimson table grape cultivar might be the most efficient producer of biomass it does not necessarily mean it produces the highest yields with the least amount of water. Having access to actual grape production data, will allow the estimation of the water productivity in terms of yields.

Conclusive remarks

In this article the Hexriver area table grape cultivation has been evaluated for the 2011/12 and the 2012/13 seasons. The article gives an indication of the versatility and applicability of the FruitLook data in providing information on water use for larger areas. Of course many other statistical methods and ways of visualisation can be adopted to gain information from the FruitLook dataset. This can include, but is certainly not limited to, linking crop growth/water use to soil type, slope aspect and weather conditions.

The new Fruitlook 2014/15 season starts on 1 October on For more information please visit the website or contact us at

Ruben Goudriaan1, Caren Jarmain2 & Andre Roux3

1 eLEAF,

2 University of KwaZulu Natal,

3 Department of Agriculture,

You may like to read these:

Go Back