This study presents a comprehensive evaluation of seasonal, locational, and varietal variations in canopy reflectance responses in 315 commercial citrus blocks from three major growing regions in Australia. The dataset includes three different citrus types (Mandarin, Navel, Valencia) and 26 varieties. The aim is to utilize this combined information to better understand yield variation and develop improved forecasting models. Landsat satellite data spanning from October 2006 to February 2021 (1419 tiles) were used to derive reflectance values, and calculate four vegetation indices (NDVI, GNDVI, LSWI, and GCVI), for each citrus block. These indices were then analyzed alongside corresponding yield data, which consisted of 3660 individual yield records dating back to 2007. Two temporal resolutions were incorporated as predictors: spatio-temporal vegetation index time series (TS) aggregated every two months and annual time series of historical block-yield records. Six statistical and machine learning algorithms were calibrated using a leave-one-year-out cross-validation approach (LOYO CV) and validated for one-year forward prediction over a five-year period (2017–2021). The results highlight significant yield variations across years, alternate bearing patterns, and spatio-temporal changes in reflectance profiles influenced by seasonal conditions, varietal characteristics, and locations. The support vector machine (SVM) algorithm with a radial basis function kernel consistently outperformed other algorithms, indicating a non-linear relationship between citrus yield and predictors. The SVM model achieved an RMSE of 15.5 T ha⁻¹ , R² of 0.88, MAE of 12.1 T ha⁻¹ , and MAPE of 29% in predicting block-yield across farms, varieties, and seasons. These prediction accuracy metrics demonstrate an improvement over current forecasting methods. Notably, the proposed approach utilizes freely available imagery, provides forecasts between two to nine months before harvest, and eliminates the need for infield counting of fruit load for image calibration. This approach provides an improved method for understanding seasonal yield variation and quantifying citrus block-yield, offering valuable insights for growers in harvest logistics, labor allocation, and resource management.

Daily calibrated multispectral imagery (Planet Fusion) of an olive irrigation deficit trial was used to assess the degree and speed to which vegetation indices indicate water stress. We developed normalization techniques to increase sensitivity to differences across a grove. The normalized difference vegetation index (NDVI) was able to significantly detect differences between the control and deficit treatments for the Arbequina variety. For the Picual variety, the green red vegetation index (GRVI) was the best indicator. Though multispectral imagery is not as quick at indicating irrigation deficits as in-field sensor data, it is complementary in being able to capture the spatial variability of water stress.

Early tree crop yield forecasts are valuable to industry and to growers, as they inform improved harvest logistics, forward selling, insurance and marketing strategies. Previous work has demonstrated the utility of weather and particularly remote sensing data to forecast tree crop yield at the orchard block scale. In this work, such data were aggregated spatially to block boundaries, and temporally at quarterly intervals. Yield prediction models were trained with a large set of grower-supplied yield data (more than 10 years, 20 orchards, 200 blocks across the Australian growing regions, for a total of 1156 yield records). Yields were forecast three months before harvest begins, and were compared to actual yields. Errors were typically around 10% and 23% at the regional and block levels respectively. Errors in 2020 were higher in non-irrigated regions due to an extreme drought in east Australia. Models were able to describe much of the variability of yields even for orchards not included in the training data, but block-level prediction errors increased by 4.1% in this case. Bootstrap sampling was used to investigate data requirements. At least 400-500 training data points was needed to minimize prediction errors. Weather data alone did not produce satisfactory accuracy, fusing weather and remote sensing data produced the best results. Including predictor data from all 8 quarterly periods from the 2 years before harvest proved a good strategy. These results demonstrate the potential of tree crop forecasting using public spatio-temporal datasets, give guidance on data requirements and identify areas for further work.

Although herbicide drifts are known worldwide and recognized as one of the major risks for crop security in the agriculture sector, the traditional assessment of damage in cotton crops caused by herbicide drifts has several limitations. The aim of this study was to assess proximal sensor and modelling techniques in the detection of phenoxy herbicide dosage in cotton crops. In situ hyperspectral data (400-900 nm) were collected at four different times after ground-based spraying of cotton crops in a factorial randomized complete block experimental design with dose and timing of exposure as factors. Three chemical doses: nil, 5% and 50% of the recommended label rate of the herbicide 2,4-D were applied to cotton plants at specific growth stages (i.e. 4-5 nodes, 7-8 nodes and 11-12 nodes). Results have shown that yield had a significant correlation (p-values <0.05) to the green peak (~550 nm) and NIR range, as the pigment and cell internal structure of the plants are key for the assessment of damage. Prediction models integrating raw spectral data for the prediction of dose have performed well with classification accuracy higher than 80% in most cases. Visible and NIR range were significant in the classification. However, the inclusion of the green band (around 550 nm) increased the classification accuracy by more than 25%. This study shows that hyperspectral sensing has the potential to improve the traditional methods of assessing herbicide drift damage.

Proximal and remote sensors have proved their effectiveness for the estimation of several biophysical and biochemical variables, including yield, in many different crops. Evaluation of their accuracy in vegetable crops is limited. This study explored the accuracy of proximal hyperspectral and satellite multispectral sensors (Sentinel-2 and WorldView-3) for the prediction of carrot root yield across three growing regions featuring different cropping configurations, seasons and soil conditions. Above ground biomass (AGB), canopy reflectance measurements and corresponding yield measures were collected from 414 sample sites in 24 fields in Western Australia (WA), Queensland (Qld) and Tasmania (Tas), Australia. The optimal sensor (hyperspectral or multispectral) was identified by the highest overall coefficient of determination between yield and different vegetation indices (VIs) whilst linear and non-linear models were tested to determine the best VIs and the impact of the spatial resolution. The optimal regression fit per region was used to extrapolate the point source measurements to all pixels in each sampled crop to produce a forecasted yield map and estimate average carrot root yield (t/ha) at the crop level. The latter were compared to commercial carrot root yield (t/ha) obtained from the growers to determine the accuracy of prediction. The measured yield varied from 17 to 113 t/ha across all crops, with forecasts of average yield achieving overall accuracies (% error) of 9.2% in WA, 10.2% in Qld and 12.7% in Tas. VIs derived from hyperspectral sensors produced poorer yield correlation coefficients (R² < 0.1) than similar measures from the multispectral sensors (R² < 0.57, p < 0.05). Increasing the spatial resolution from 10 to 1.2 m improved the regression performance by 69%. It is impossible to non-destructively estimate the pre-harvest spatial yield variability of root vegetables such as carrots. Hence, this method of yield forecasting offers great benefit for managing harvest logistics and forward selling decisions.

While studies on deforestation of protected areas (PAs) have been conducted in many parts of the world, no comparative study has been done over an entire country in the tropics. Thus, we conducted a country-wide assessment of forest cover loss in all terrestrial protected areas of the Philippines, covering 198 PAs with a total area of 4.68 million ha. This study utilised Hansen's Landsat-derived global maps of forest cover change from 2000 to 2012, with tree canopy cover data for 2000 as the base year. Correlation and logistic regression analyses were employed to determine the significance and magnitude of the relationships between forest cover and 11 predictor variables. The assessment of forest loss reveals that the terrestrial protected areas are generally effective in reducing forest loss. Over the 12-year period, the average rate (2.59%) of forest clearing in protected areas is marginally lower by 0.1% than the entire country (2.69%). Within the same duration, the average forest loss rate within the 2-km buffer zones of selected protected areas is 1.4 times of those inside PAs. However, there was a significant number of PAs with phenomenal forest cover loss in terms of extent (48,583 ha over 12 years) and rate (up to 21%). We found that spatial predictor variables included in this study have weak or no relationships with forest cover, and hence they are not reliable inputs for predictive modelling. Comprehensive assessments of deforestation are needed at the micro-scale (e.g. single PA level) level and relatively shorter historical timeframe (e.g. less than a decade), to generate useful information for policy formulation, planning, and management.

The aim of this study was to develop a method that will use satellite imagery to identify areas of high forest growth and productivity, as a primary input in prioritising revegetation sites for carbon sequestration. Using the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data, this study analysed the annual net primary production (NPP) values (gC/m²) of images acquired from 2000 to 2013, covering the Condamine Catchment in southeast Queensland, Australia. With the analysis of annual rainfall data during the same period, three transitions of "normal to dry" years were identified to represent the future climate scenario considered in this study. The difference in the corresponding NPP values for each year was calculated, and subsequently averaged to the get the "Mean of Annual NPP Difference" (MAND) map. This layer identified the areas with increased net primary production despite the drought condition in those years. Combined with key thematic maps (i.e. regional ecosystems, land use, and tree canopy cover), the priority areas were mapped. The results have shown that there are over 42 regional ecosystem (RE) types in the study area that exhibited positive vegetation growth and productivity despite the decrease in annual rainfall. However, seven (7) of these RE types represents the majority (79 %) of the total high productivity area. A total of 10,736 ha were mapped as priority revegetation areas. This study demonstrated the use of MODIS-NPP imagery to map vegetation with high carbon sequestration rates necessary in prioritising revegetation sites.

Yield loss in crops is often associated with plant disease or external factors such as environment, water supply and nutrient availability. Improper agricultural practices can also introduce risks into the equation. Herbicide drift can be a combination of improper practices and environmental conditions which can create a potential yield loss. As traditional assessment of plant damage is often imprecise and time consuming, the ability of remote and proximal sensing techniques to monitor various bio-chemical alterations in the plant may offer a faster, non-destructive and reliable approach to predict yield loss caused by herbicide drift. This paper examines the prediction capabilities of partial least squares regression (PLSR) models for estimating yield. Models were constructed with hyperspectral data of a cotton crop sprayed with three simulated doses of the phenoxy herbicide 2,4-D at three different growth stages. Fibre quality, photosynthesis, conductance, and two main hormones, indole acetic acid (IAA) and abscisic acid (ABA) were also analysed. Except for fibre quality and ABA, Spearman correlations have shown that these variables were highly affected by the chemical. Four PLS-R models for predicting yield were developed according to four timings of data collection: 2, 7, 14 and 28 days after the exposure (DAE). As indicated by the model performance, the analysis revealed that 7 DAE was the best time for data collection purposes (RMSEP = 2.6 and R² = 0.88), followed by 28 DAE (RMSEP = 3.2 and R² = 0.84). In summary, the results of this study show that it is possible to accurately predict yield after a simulated herbicide drift of 2,4-D on a cotton crop, through the analysis of hyperspectral data, thereby providing a reliable, effective and non-destructive alternative based on the internal response of the cotton leaves.

Climate and agriculture affect each other in a reciprocal fashion. In agrarian countries like Bangladesh, agricultural activities are mostly defined by seasonal climatic cycles. Failure of agricultural adaptations to keep pace with climate change and variabilities have the potential to impact food production, and eventually, food security. Exploration of agricultural impacts of climate change in coastal areas of Bangladesh, one of the globally top ranked climate vulnerable countries, was the guiding focus of this study. Literature review, farmer interviews, agricultural office visits, government organizational databases, global climate model ensembles, and tide gauge records were used to collect primary and secondary information. Field data was collected from randomly selected 381 farmers from 10 selected subdistricts across the coastal areas of Bangladesh during September–October 2018. A wide range of statistical and econometric approaches were applied to reveal the complex relationship between farmers’ perception, climatic data, farming variables and socioeconomic characteristics.

Farming decisions in relation to adaptations under climate change largely depend on perception of climate change, and their feasibility is linked to the accuracy of their perceptions. Last 30- year (1988–2017) average temperature shows 0.45 °C spatial differences among the visited subdistricts. Yearly precipitation gradient could be >100 cm from the drier western to the wetter eastern coasts. While monthly averages of coastal temperature had increased except in early winter (October–December), while pre-monsoon and November rainfall had decreased with an increase in monsoon precipitation. Onset of monsoon rainfall was found to be delayed in the coastal areas. The farmers, in general, mentioned a warmer temperature and less rainfall in the recent decade (2009–2018) compared with the past decade (1999–2008). Their perceptions were mostly consistent with meteorological records though the observed decrease in winter temperature and the change in rainfall in some locations did not match with their perceptions. About one-third (30%) of the farmers accurately identified the changes in annual rainfall and temperature (annual, summer and winter average). Cluster analysis flagged 58.8% of the farmers as weak perception group. However, 41.2% of them were found in the moderate perception group characterized by younger age, better education, smaller family size, richer economic status, larger farm size, more affiliation with non-farm jobs, users of more communication media, closer to the marketplaces, and more distant from the sea. Thus, they were comparatively economically better-off than the weak perception group.

Farm productivity had a mean value of 1.98 in terms of revenue-cost ratio as reported by the farmers during the interviews based on the previous cropping year. Over one in ten (11%) of the farmers opined that their farm productivity had currently declined compared with the past. Majority (64%) of the farmers thought that this decline was due to climate change and its consequences, such as changes in temperature, precipitation, floods, droughts, and salinity. Outputs of the logistic regression shows that the farmers with greater level of education, more awareness of climate change, less communication with extension agents, stronger belief in decreased cyclone and salinity, and weaker belief in decreased flood had perceived that climate change was responsible for the decrease in their farm productivity. The farmers identified dry season soil salinity, coastal inundations and floods were the climate change induced issues that had adversely affected crop productivity.

To keep the farm productivity at desired levels, the farmers had adopted on average 10–11 farm management practices out of the 22 selected adaptation options. Two-thirds (67%) of the farmers mentioned that they had changed the farm management practices because of climate change. The farmers performed the crop-related adaptations more than the livestock, fisheries or general agricultural adaptations. According to the discriminant function analysis, the farmers with stronger belief in climatic impacts on their farm management were younger in age, had higher level of education, more involvement with non-farm jobs, greater affiliation with farmrelated organizations, more awareness of climate change, and greater accuracy of perception of changes in climatic variables.

Similar to the changes in farm management practices, 64% of the farmers had changed their farming systems due to climate change. In recent years (2009–2018) compared with the previous years (1999–2008), three farm enterprises, namely rice, vegetables, and livestock, had decreased, while three others, namely fisheries, forestry, and fruit farming, had increased. The random forest algorithm has identified that larger family size had negative effect, while age, education, and cultivated land had positive effects on the probability of believing that climate change had impacted their changes in farming systems. The farmers who more accurately perceived the changes in temperature, rainfall and cyclones and had better awareness of climate change agreed in greater magnitude that their farming systems had changed due to the influence of climate change.

Farmer opinions highlight the adverse effects of salinity intrusion, temperature, rainfall, floods, and cyclones on their agricultural activities. Therefore, we modelled coastal inundation in Bangladesh by semi-empirical approach using downscaled and bias corrected 28 global climate models. Singular spectrum analysis was undertaken to separate the trends from the time series of temperature and tide gauge data for the period of 1980–2100. The model shows that sea level is likely to rise at a rate of 6.69–9.88 mm/year which would result in up to 1.15 m sealevel rise by 2100 inundating at least 2098 km2 of the coast which has =1 m elevation. Though this inundated part is located mostly outside the river and coastal embankments, saline water intrusion and groundwater contamination are likely to increase in a changing climate.

During historical disasters (ten selected flood and cyclone events) between 1970 and 2017, coastal areas had lost 12.10% of crop production which was 2.54% higher than the non-coast. Temperature and precipitation affect crop yield and production in two ways—through their trends and variabilities. The mixed effects model reveals that these variables explain 12% of the variance in crop production. Climate trends and variabilities are likely to reduce crop yield, respectively, by 2.75 and 2.91%, which equates to 2.4 million metric tons of crop loss per year.

Farmers’ concerns and data-driven analysis establish the fact that coastal agriculture is increasingly under climatic threats and in more precarious conditions than the inland agriculture. Climatic impacts on coastal farming cannot be stopped but there is scope to keep it viable under climate change. This study suggests that economically worse-off farmers should be attended and communicated with updated climate change information by extension agents to enhance their adaptation actions. Failure of around one-third of the farmers to detect the climatic impacts on farm productivity, farm management and farming systems implies that they need to be updated with farm-related climatic knowledge to motivate them to adopt agricultural adaptations. Enhancing involvement of the farmers with agricultural extension associations is likely to improve their climate change awareness and perception of changes in climatic variables. Limited capacity of the farmers to keep the coastal farming sustainable warrants for external support. For example, maintenance of river and coastal embankments should continue to be the first priority of coastal agricultural planning. This research provides information and insights of climatic impacts on coastal farms from both farmer and empirical perspectives, which are necessary for agricultural policy formulation. Information generated through this study is expected to help policymakers and extension agents to formulate and implement coastal agricultural development programmes in Bangladesh. Researchers and academicians could benefit from the approaches and methods used here to apply in various socioeconomic and ecological constellations.