Crop yield prediction from UAV images has significant potential in accelerating and revolutionizing crop breeding pipelines. Although convolutional neural networks (CNN) provide easy, accurate and efficient solutions over traditional machine learning models in computer vision applications, a CNN training requires large number of ground truth data, which is often difficult to collect in the agricultural context. The major objective of this study was to develope an end-to-end 3D CNN model for plot-scale soybean yield prediction using multitemporal UAV-based RGB images with approximately 30,000 sample plots. A low-cost UAV-RGB system was utilized and multitemporal images from 13 different experimental fields were collected at Argentina in 2021. Three commonly used 2D CNN architectures (i.e., VGG, ResNet and DenseNet) were transformed into 3D variants to incorporate the temporal data as the third dimension. Additionally, multiple spatiotemporal resolutions were considered as data input and the CNN architectures were trained with different combinations of input shapes. The results reveal that: (a) DenseNet provided the most efficient result (R² 0.69) in terms of accuracy and model complexity, followed by VGG (R² 0.70) and ResNet (R² 0.65); (b) Finer spatiotemporal resolution did not necessarily improve the model performance but increased the model complexity, while the coarser resolution achieved comparable results; and (c) DenseNet showed lower clustering patterns in its prediction maps compared to the other models. This study clearly identifies that multitemporal observation with UAV-based RGB images provides enough information for the 3D CNN architectures to accurately estimate soybean yield non-destructively and efficiently.

Human-induced climate change significantly alters the spatiotemporal characteristics of climate zones, which drives agricultural land use and ecosystem change. However, the detectability of shifting climate zones and the rate and time of the changes has yet to be adequately addressed at the regional-to-local scale. We mapped and analyzed changes to temperature and precipitation across Kenya during the past four decades, and linked those changes to shifts in the geographic distribution and arrangement of climate zones at regional scales. We observed an approximate 1 °C increase in average annual temperature over the 40-year period. A total of 76,346 km² shifted from cooler to hotter zones, while 1298 km² shifted from hotter to cooler zones. Tropical climate regions expanded from 91 to 93%, with over 13,000 km² shifting from alpine and temperate to tropical regions. Average annual precipitation demonstrated little or no trend, but substantial spatial changes were observed. A total of 136,129 km² shifted from wetter to drier zones, while 23,317 km² shifted from drier to wetter zones. Arid climate regions expanded from 72 to 81%, a roughly 50,000 km² shift from humid and semi-humid-to-semi-arid to arid regions. Overall, there was a 207,557 km² shift in temperature and precipitation zones. As the climate zones predominately shift toward hotter and drier conditions, climatic diversity will decline, and in turn, ecosystem diversity and the ecosystem goods and services to society will decline. The changes also have broader global implications in terms of their contribution to global drylands as well as influencing earth system cycles. Overall, such information can better inform the Kenyan National Climate Change Response Strategy and be used to reach the UN Sustainable Development Goals.

Remote sensing data are increasingly being used in digital archaeology for the potential non-invasive detection of archaeological remains. The purpose of this research is to evaluate the capability of standalone (LiDAR and aerial photogrammetry) and integration/fusion remote sensing approaches in improving the prospecting and interpretation of archaeological remains in Cahokia’s Grand Plaza. Cahokia Mounds is an ancient area; it was the largest settlement of the Mississippian culture located in southwestern Illinois, USA. There are a limited number of studies combining LiDAR and aerial photogrammetry to extract archaeological features. This article, therefore, combines LiDAR with photogrammetric data to create new datasets and investigate whether the new data can enhance the detection of archaeological/ demolished structures in comparison to the standalone approaches. The investigations are implemented based on the hillshade, gradient, and sky view factor visual analysis techniques, which have various merits in revealing topographic features. The outcomes of this research illustrate that combining data derived from different sources can not only confirm the detection of remains but can also reveal more remains than standalone approaches. This study demonstrates that the use of combination remote sensing approaches provides archaeologists with another powerful tool for site analysis.

Archaeological feature detection at complex, mixed-environment sites requires accurate, efficient methods for identifying subtle morphological signatures. This study presents a semi-automated remote sensing pipeline for the detection and delineation of archaeological earthworks at Cahokia Mounds (Illinois, USA), a major Mississippian urban center and UNESCO World Heritage Site. Three LiDAR datasets, two collected via UAV-mounted sensors and one from a piloted aircraft survey, were processed into Digital Terrain Models and transformed into Local Relief Models (LRM). K-means clustering was applied to segment the LRMs into feature classes, followed by contour bounding using the OpenCV library to outline mounds and borrow pits. Additionally, SAR-derived Local Incidence Angle (LIA) rasters from PALSAR-3 and Sentinel-1 were processed through angular deviation mapping to identify slope anomalies associated with archaeological features. Results across all five datasets demonstrate the complementary strengths of LiDAR and SAR: LiDAR excels at resolving elevation-defined features such as mound footprints, while LIA captures directional slope behavior that highlights mound edges, borrow pit rims, and linear features such as causeways. Comparative analysis of LiDAR acquisition frequencies reveals minimal differences in archaeological feature recovery between pulse settings, suggesting that sensor platform choice matters more than power-density tradeoffs for this application. Despite the need for human review to filter modern disturbances and natural false positives, the integrated workflow meaningfully accelerates prospection and reduces interpretive subjectivity. The methods are scalable, site-invariant, and work with open-access data, making them applicable to archaeological landscapes worldwide.