Energy balance and irrigation performance assessments in lemon orchards by applying the SAFER algorithm to Landsat 8 images

In the semiarid conditions of the São Francisco River Basin, Brazil, irrigated fruit crops are replacing natural vegetation, being lemon highlighted by incentives for the national and international markets. This paper aimed to support the rational water management of lemon orchards under different irrigation systems under these conditions. We present a methodology based on the use of the visible and near infrared images from Landsat 8 (L8) satellite together with weather and actual yield data (Ya), to assess the energy balance components and irrigation performance indicators (IPI) by applying the SAFER (Simple Algorithm for Evapotranspiration Retrieving) in six commercial farms under different irrigation systems inside the basin in the northern Minas Gerais state, Southeast Brazil. The ET rates averaged 2.7 mm d−1, 2.9 mm d−1, and 3.7 mm d−1, for drip, micro sprinkler, and pivot irrigated orchards, respectively. The evaporative fraction (latent heat flux by the available energy) reached above 1.00 for localized irrigation (drip and micro sprinkler), and 1.30 for pivots, during the lemon phenological stages from fruit growth to harvest peaks. Pivot irrigation systems were not recommended under the semi-arid conditions, due to large water direct evaporated from the soil and air close to the surface. For drip and micro sprinkler irrigated orchards, crop coefficient curves were modeled based on the accumulated degree-days (DDac) to estimate lemon crop water requirements. Drip irrigated orchards presented better water productivity levels being recommended together with deficit irrigation strategies which could allow good lemon yields with saving water savings. The most important findings of the current research are that the SAFER algorithm can be applied to estimate crop ET with satellite images without the thermal bands which together with modelled Ya data, irrigation assessments can be carried out at high spatial resolution following the principles of precision agriculture. For replication of the methods in other regions, simple calibrations of the modelling equations can be performed to infer the specific environmental conditions.