Pre-harvest yield estimation of mango fruit is important for the optimization of inputs and other resources on the farm. Current industry practice of visual counting the fruit on a small number of trees for yield forecasting can be highly inaccurate due to the spatial variability, especially if the trees selected do not represent the entire crop. Therefore, this study evaluated the potential of high resolution WorldView-3 (WV3) satellite imagery to estimate yield of mango by integrating both geometric (tree crown area) and optical (spectral vegetation indices) data using artificial neural network (ANN) model. WV3 images were acquired in 2016–2017 and 2017–2018 growing seasons at the early fruit stage from three orchards in Acacia Hills region, Northern Territory, Australia. Stratified sampling technique (SST) was applied to select 18 trees from each orchard and subsequently ground truthed for yield (kg·tree⁻¹) and fruit number per tree. For each sampled tree, spectral reflectance data and tree crown area (TCA) was extracted from WV3 imagery. The TCA was identified as the most important predictor of both fruit yield (kg·tree⁻¹) and fruit number, followed by NDVI red-edge band when all trees from three orchards in two growing seasons were combined. The results of all sampled trees from three orchards in two growing seasons using ANN model produced a strong correlation (R² = 0.70 and 0.68 for total fruit yield (kg·tree⁻¹) and fruit number respectively), which suggest that the model can be obtained to predict yield on a regional level. On orchard level also the ANN model produced a high correlation when both growing seasons were combined. However, the model developed in one season could not be applied in another season due to the influence of seasonal variation and canopy condition. Using the relationship derived from the measured yield parameters against combined VIs and TCA data, the total fruit yield (t·ha⁻¹) and fruit number were estimated for each orchard, produced 7% under estimation to less than 1% over estimation. The accuracy of the findings showed the potential of WV3 imagery to better predict the yield parameters than the current practice across the mango industry as well as to quantify lost yield as a result of delayed harvest.

Accurate pre-harvest estimation of avocado (Persea americana cv. Haas) yield offers a range of benefits to industry and growers. Currently there is no commercial yield monitor available for avocado tree crops and the manual count method used for yield forecasting can be highly inaccurate. Remote sensing using satellite imagery offers a potential means to achieve accurate pre-harvest yield forecasting. This study evaluated the accuracies of high resolution WorldView (WV) 2 and 3 satellite imagery and targeted field sampling for the pre-harvest prediction of total fruit weight (kg·tree⁻¹) and average fruit size (g) and for mapping the spatial distribution of these yield parameters across the orchard block. WV 2 satellite imagery was acquired over two avocado orchards during 2014, and WV3 imagery was acquired in 2016 and 2017 over these same two orchards plus an additional three orchards. Sample trees representing high, medium and low vigour zones were selected from normalised difference vegetation index (NDVI) derived from the WV images and sampled for total fruit weight (kg·tree⁻¹) and average fruit size (g) per tree. For each sample tree, spectral reflectance data was extracted from the eight band multispectral WV imagery and 18 vegetation indices (VIs) derived. Principal component analysis (PCA) and non-linear regression analysis was applied to each of the derived VIs to determine the index with the strongest relationship to the measured total fruit weight and average fruit size. For all trees measured over the three year period (2014, 2016, and 2017) a consistent positive relationship was identified between the VI using near infrared band one and the red edge band (RENDVI1) to both total fruit weight (kg·tree⁻¹) (R² = 0.45, 0.28, and 0.29 respectively) and average fruit size (g) (R² = 0.56, 0.37, and 0.29 respectively) across all orchard blocks. Separate analysis of each orchard block produced higher R² values as well as identifying different optimal VIs for each orchard block and year. This suggests orchard location and growing season are influencing the relationship of spectral reflectance to total fruit weight and average fruit size. Classified maps of avocado yield (kg·tree⁻¹) and average fruit size per tree (g) were produced using the relationships developed for each orchard block. Using the relationships derived between the measured yield parameters and the optimal VIs, total fruit yield (kg) was calculated for each of the five sampled blocks for the 2016 and 2017 seasons and compared to actual yield at time of harvest and pre-season grower estimates. Prediction accuracies achieved for each block far exceeded those provided by the grower estimates.

Early prediction of sugarcane crop yield at the commercial block level (unit area of a single crop of the same variety, ratoon or planting date) offers significant benefit to growers, consultants, millers, policy makers, crop insurance companies and researchers. This current study explored a remote sensing based approach for predicting sugarcane yield at the block level by further developing a regionally specific Landsat time series model and including individual crop sowing (or previous seasons’ harvest) date. For the Bundaberg growing region of Australia this extends over a five months period, from July to November. For this analysis, the sugarcane blocks were clustered into 10 groups based on their specific planting or ratoon commencement date within the specified five months period. These clustered or groups of blocks were named ‘bins’. Cloud free (<20%) satellite data from the polar-orbiting Landsat-8 (launched 2013), Sentinel-2A (launched 2015) and Sentinel-2B (launched 2017) sensors were acquired over the cane growing region in Bundaberg (area of 32,983 ha), from the growing season starting in July 2014, with the average green normalised difference vegetation index (GNDVI) derived for each block. The number of images acquired for each season was defined by the number of cloud free acquisitions. Using the Simple Linear Machine Learning (ML) algorithm, the extracted Landsat derived GNDVI values for each of the blocks were converted to Sentinel GNDVI. The average GNDVI of each ‘bin’ was plotted and a quadratic model was fitted through the time series to identify the peak growth stage defined as the maximum GNDVI value. The model derived maximum GNDVI values for each of the bins were then regressed against the average actual yield (t·ha^-1) achieved for the respective bin over the five growing years, producing strong correlations (R² = 0.92 to 0.99). The quadratic curves developed for the different bins were shifted according to the specific planting or ratoon date of an individual block allowing for the peak GNDVI value of the block to be calculated, regressed against the actual block yield (t·ha^-1) and the prediction of yield to be made. To validate the accuracies of the 10 time series algorithms representing each of the 10 bins, 592 individual blocks were selected from the Bundaberg region during the 2019 harvest season. The crops were clustered into the appropriate bins with the respective algorithm applied. From a Sentinel image acquired on the 5 May 2019, the prediction accuracies were encouraging (R² = 0.87 and RMSE = 11.33 (t·ha^-1)) when compared to actual harvested yield, as reported by the mill. The results presented in this paper demonstrate significant progress in the accurate prediction of sugarcane yield at the individual sugarcane block level using a remote sensing, time-series based approach.

Accurate pre-harvest yield forecasting of mango is essential to the industry as it supports better decision making around harvesting logistics and forward selling, thus optimizing productivity and reducing food waste. Current methods for yield forecasting such as manually counting 2–3% of the orchard can be accurate but are very time inefficient and labour intensive. More recent evaluations of technological solutions such as remote (satellite) and proximal (on ground) sensing have provided very encouraging results, but they still require infield in-season sampling for calibration, the technology comes at a significant cost, and commercial availability is limited, especially for vehicle-mounted sensors. This study presents the first evaluation of a ”time series”—based remote sensing method for yield forecasting of mango, a method that does not require infield fruit counts and utilizes freely available satellite imagery. Historic yield data from 2015 to 2022 were sourced from 51 individual orchard blocks from two farms (AH and MK) in the Northern Territory of Australia. Time series measures of the canopy reflectance properties of the blocks were obtained from Landsat 7 and 8 satellite data for the 2015–2022 growing seasons. From the imagery, the following vegetation indices (VIs) were derived: EVI, GNDVI, NDVI, and LSWI, whilst corresponding weather variables (rainfall (Prec), temperature (Tmin/Tmax), evapotranspiration (ETo), solar radiation (Rad), and vapor pressure deficit (vpd)) were also sourced from SILO data. To determine the relationships among weather and remotely sensed measures of canopy throughout the growing season and the yield achieved (at the block level and the farm level), six machine learning (ML) algorithms, namely random forest (RF), support vector regression (SVR), eXtreme gradient boosting (XGBOOST), RIDGE, LASSO and partial least square regression (PLSR), were trialed. The EVI/GNDVI and Prec/Tmin were found to be the best RS and weather predictors, respectively. The block-level combined RS/weather-based RF model for 2021 produced the best result (MAE = 2.9 t/ha), marginally better than the RS only RF model (MAE = 3.4 t/ha). The farm-level model error (FLEM) was generally lower than the block-level model error, for both the combined RS/weather-based RF model (farm = 3.7%, block (NMAE) = 33.6% for 2021) and the RS-based model (farm = 4.3%, block = 38.4% for 2021). Further testing of the RS/weather-based RF models over six additional orchards (other than AH and MK) produced errors ranging between 24% and 39% from 2016 to 2020. Although accuracies of prediction did vary at both the block level and the farm level, this preliminary study demonstrates the potential of a ”time series” RS method for predicting mango yields. The benefits to the mango industry are that it utilizes freely available imagery, requires no infield calibration, and provides predictions several months before the commercial harvest. Therefore, this outcome not only presents a more adoptable option for the industry, but also better supports automation and scalability in terms of block-, farm-, regional, and national level forecasting.