This work presents quantitative detection of water stress and estimation of the water stress level: none, light, moderate, and severe on potato crops. We use hyperspectral imagery and state of the art machine learning algorithms: random decision forest, multilayer perceptron, convolutional neural networks, support vector machines, extreme gradient boost, and AdaBoost. The detection and estimation of water stress in potato crops is carried out on two different phenological stages of the plants: tubers differentiation and maximum tuberization. The machine learning algorithms are trained with a small subset of each hyperspectral image corresponding to the plant canopy. The results are improved using majority voting to classify all the canopy pixels in the hyperspectral images. The results indicate that both detection of water stress and estimation of the level of water stress can be obtained with good accuracy, improved further by majority voting. The importance of each band of the hyperspectral images in the classification of the images is assessed by random forest and extreme gradient boost, which are the machine learning algorithms that perform best overall on both phenological stages and detection and estimation of water stress in potato crops.

Potato farming is relevant for global carbon balances and greenhouse emissions, of which gross primary productivity (GPP) is one of the main drivers. In this study, the net carbon ecosystem exchange (NEE) was measured using the Eddy Covariance (EC) method in two potato crops, one of them with an irrigation system, the other under rainfed conditions. Accurate NEE partition into GPP and ecosystem respiration (RECO) was carried out by fitting a light response curve. Direct measurements of dry weight and leaf area were performed from sowing to the end of canopy life cycle and tuber bulking. Agricultural drought in the rainfed crop resulted in limited GPP rate, low leaf area index (LAI), and low canopy carbon assimilation response to the photosynthetically active radiation (PAR). Hence, in this crop, there was lower efficiency in tuber biomass gain and NEE sum indicated net carbon emissions to atmosphere (NEE = 154.7 g C m⁻² ± 30.21). In contrast, the irrigated crop showed higher GPP rate and acted as a carbon sink (NEE = −366.6 g C m⁻² ± 50.30). Our results show, the environmental and productive benefits of potato crops grown under optimal water supply.

Gross primary productivity (GPP) is an essential parameter to assess the efficiency of terrestrial ecosystems on carbon transfer. Although GPP is regularly measured with eddy covariance (EC) systems, these are restricted to the tower footprint area, and remote sensing (RS) products have estimated GPP using multispectral vegetation indexes (VIs) from farms to whole ecosystems. Indeed, nowadays, unmanned aerial vehicle (UAV)-based RS technology is becoming more accessible. Accordingly, we propose the estimation of GPP using VIs at high spatial resolutions using UAVs and multi-spectral cameras. A small typical farm in Colombia was cultivated with carrot as our base crop. An EC system was installed to estimate GPP and was used as a reference. A total of 24 VIs from UAV-based RS products were selected and compared with the GPP of the EC system. A cross-validation process was performed, and seven VIs obtained a high R² score (0.76–0.78). The accumulated GPP, estimated with the best index (NIRv) was 520.3 g C m⁻², while the GPP-EC estimate was 580.4 g C m⁻² (10.3% error). This work showed that it is possible to estimate the GPP of carrot crops using UAV-based RS, VIs, and linear regression models, which can be used in further research on GPP using UAVs.

Estimating gross primary production (GPP) is important to understand the land–atmosphere CO₂ exchange for major agroecosystems. Eddy covariance (EC) measurements provide accurate and reliable information about GPP, but flux measurements are often not available. Upscaling strategies gain importance as an alternative to the limitations of the use of the EC. Although the potato provides an important agroecosystem for worldwide carbon balance, there are currently no studies on potato GPP upscaling processes. This study reports two GPP scaling-up approaches from the detailed leaf-level characterization of gas exchange of potatoes. Multilayer and big leaf approaches were applied for extrapolating chamber and biometric measurements from leaf to canopy. Measurements of leaf area index and photosynthesis were performed from planting to the end of the canopy life cycle using an LP-80 ceptometer and an IRGA Li-Cor 6800, respectively. The results were compared to concurrent measurements of surface–atmosphere GPP from the EC measurements. Big-leaf models were able to simulate the general trend of GPP during the growth cycle, but they overestimated the GPP during the maximum LAI phase. Multilayer models correctly reproduced the behavior of potato GPP and closely predicted both: the daily magnitude and half-hourly variation in GPP when compared to EC measurements. Upscaling is a reliable alternative, but a good treatment of LAI and the photosynthetic light-response curves are decisive factors to achieve better GPP estimates. The results improved the knowledge of the biophysical control in the carbon fluxes of the potato crop.