Irrigation is commonly viewed as an activity fixed in time and place requiring permanent infrastructure. However, smallholder farmers in Sub-Saharan Africa engage in irrigation in diverse locations under different organizational modalities. This research analyses a flexible and dynamic form of irrigation driven by unique partnerships between migrants and local actors who derive benefits from land and water resources along a sand river. The case study of the ephemeral Olkeriai sand river in Kajiado, Kenya, was based on a baseline survey of 107 farm plots and 23 in-depth interviews with farmers. We found that 75% of the farm plots were managed under transient farming partnerships between migrant farmers and capital providers, locally known as tajiris, who leased land from local landowners to grow high-value market crops. These partnerships are based on flexible agreements between the actors and the frequent need to review and re-build them creates a fertile ground for new entrepreneurial players from within and outside the area, playing a key role in accelerating irrigation intensification and expansion. However, these irrigation ventures are not always successful. Unreliable partners, unstable market channels and increasing costs of irrigation inputs frequently result in farm losses and breaking up of partnership ventures. Furthermore, the lack of collective action among various resource users at the catchment level raises questions on how to sustainably manage the natural resources in sand rivers.

The growing competition for finite land and water resources and the need to feed an ever-growing population require new techniques to monitor the performance of irrigation schemes and improve land and water productivity. Datasets from FAO's portal to monitor Water Productivity through Open access Remotely sensed derived data (WaPOR) are increasingly applied as a cost-effective means to support irrigation performance assessment and identify possible pathways for improvement. This study presents a framework that applies WaPOR data to assess irrigation performance indicators, including uniformity, equity, adequacy, and land and water productivity differentiated by irrigation method (furrow, sprinkler, and centre pivot) at the Xinavane sugarcane estate, Mozambique. The WaPOR data on water, land, and climate are in near-real time and spatially distributed, with the finest spatial resolution in the area of 100gm. The WaPOR data were first validated agronomically by examining the biomass response to water, and then the data were used to systematically analyse seasonal indicators for the period 2015 to 2018 on g1/48000gha. The WaPOR-based yield estimates were found to be comparable to the estate-measured yields with ±20g% difference, a root mean square error of 19±2.5gtgha-1 and a mean absolute error of 15±1.6gtgha-1. A climate normalization factor that enables the spatial and temporal comparison of performance indicators are applied. The assessment highlights that in Xinavane no single irrigation method performs the best across all performance indicators. Centre pivot compared to sprinkler and furrow irrigation shows higher adequacy, equity, and land productivity but lower water productivity. The three irrigation methods have excellent uniformity (g1/494g%) in the four seasons and acceptable adequacy for most periods of the season except in 2016, when a drought was observed. While this study is done for sugarcane in one irrigation scheme, the approach can be broadened to compare other crops across fields or irrigation schemes across Africa with diverse management units in the different agroclimatic zones within FAO WaPOR coverage. We conclude that the framework is useful for assessing irrigation performance using the WaPOR dataset.

Efficient irrigation water management for an 18,000 ha sugarcane plantation in Xinavane in southern Mozambique is a challenge. Sugarcane is an irrigation intensive crop and its productivity is sensitive to water stress. Options to adopt field water management best practices and proper irrigation scheduling are limited due to the lack of plot-level information on the actual crop water use and stress levels throughout the growing season. Due to heterogeneity in cropping calendar within the sugarcane plantation, at a certain point of time, different plots are at different growth stages. This makes scheme level irrigation scheduling complex and calls for frequent crop water use information. To fill this gap, this study presents a novel approach where a combination of satellite imagery with local weather data is used to provide daily evaporation rates. The Priestley–Taylor equation is applied to quantify evaporation (soil evaporation + transpiration) using radiation and temperature data from a meteorological station and spatial albedo estimates derived from the Sentinel-2 satellites. The results show 20 meter resolution maximum crop evaporation estimates can be derived with the proposed methodology. Additionally, the results show NDVI in the last two crop stages is able to distinguish between poor and good performing fields. Therefore, NDVI can be a useful index to estimate actual evaporation. First, the evaporation estimates were corrected for the crop stage using NDVI proxies and an additional stress indicator was used to calculate the actual evaporation flux spatially. The spatial evaporation estimates provide the water manager with information on actual crop water use and biomass development, which is relevant to both crop monitoring and irrigation management water management when drought-related stress is filtered.

Crop water productivity (CWP) has become a recognised indicator in assessing the state of Sustainable Development Goals (SDG) 6.4—to substantially increase water use efficiency. This indicator, while useful at a global scale, is not comprehensive at a local scale. To fill this gap, this research proposes a CWP framework, that takes advantage of the spatio-temporal availability of remote sensing, that identifies CWP goals and sub-indicators specific to the needs of the targeted domain. Three sub-indicators are considered; (i) a global water productivity score (GWPS), (ii) a local water productivity score (LWPS) and (iii) a land and water use productivity score (YWPS). The GWPS places local CWP in the global context and focuses on maximised CWP. The LWPS differentiates yield zones, normalising for potential product, and focuses on minimising water consumption. The YWPS focuses simultaneously on improving land and water productivity equally. The CWP framework was applied to potato in the West Bank, Palestine. Three management practices were compared under each sub-indicator. The case study showed that fields with high and low performance were different under each sub-indicator. The performance associated with different management practices was also different under each sub-indicator. For example, a winter rotation had a higher performance under the YWPS, the fall rotation had a higher performance under the LWPS and under the GWPS there was little difference. The results showed, that depending on the basin goal, not only do the sub-indicators required change, but also the management practices or approach required to reach those basin goals. This highlights the importance of providing a CWP framework with multiple sub-indicators, suitable to basin needs, to ensure that meeting the SDG 6.4 goal does not jeopardise local objectives.