Combating Climate Change: SC Solutions’ Remote Sensing Innovation
Vegetation Drought in the Face of Climate Change
As our planet confronts the multifaceted challenges of climate change, scientists, agricultural experts, and environmentalists are exploring innovative solutions to mitigate its effects. Traditional methods such as fertilizers and pesticides, while initially effective, often lose their efficacy over time. Genetic engineering, though promising, typically requires extensive periods of development and implementation. Thus, alternative solutions are imperative.
SC Solutions is at the forefront of utilizing remote sensing technology to identify genetically robust vegetation. These hardy plants can be selected and propagated to better withstand the adverse effects of climate change. This article delves into our advanced methodologies.
Figure 1: Robust vegetation in desert area
Background ━ Vegetation Monitoring via Remote Sensing
Remote sensing offers several sophisticated methods for monitoring vegetation health and resilience. Among these methods are various optical spectral indices, which are critical for assessing plant vigor.
The figure below of London Hyde Park serves as an example. The left image represents natural RGB colors as perceived by the human eye, while the right image depicts the intensity of near-infrared light captured by the Sentinel-2 satellite (Sentinel-2 B8A). This contrast clearly illustrates the amount of infrared detected by the satellite. Healthy plants with higher chlorophyll content reflect more near-infrared energy compared to unhealthy plants.
Figure 2: Contrast between an RGB satellite image and an infrared satellite image
A good way to analyze vegetation is via optical indices. These indices are constructed using various spectral intensities, by comparing their relative intensities or with more complicated mathematical operations.
Some examples include:
Normalized Difference Vegetation Index (NDVI)
The NDVI compares the intensity between reflected Infrared with reflected Red spectral bands.
Figure 3: Contrast between an RGB satellite and an NDVI index image
Healthy vegetation reflected more infrared than red light.
Normalized Difference Moisture Index (NDMI)
The NDMI compares the intensity between reflected Near-Infrared with reflected Short-wavelength Infrared spectral bands.
Figure 4: Contrast between an RGB satellite image and an NDMI index image
This is because water absorbs more short-wavelength (SW) infrared light. Higher moisture levels will reduce the intensity of reflected SW infrared.
SC Solutions’ Proprietary Indices
At SC Solutions, we have developed proprietary indices tailored to specific environmental conditions. One such index is the SC Vegetation Index (SCVI), which indicates vegetation robustness against drought conditions. Unlike traditional indices such as NDVI, SCVI incorporates additional data inputs and time-series analyses to provide a comprehensive assessment. It differentiates more towards plant health and robustness than NDVI or NDMI.
Figure 5: SC-developed index compared with RGB
Botanic Analyzer ━ Big AI Model for Vegetation Analysis
Understanding vegetation performance over time is crucial for accurate assessment. For example, a time-series plot of NDRE values for two corn varieties can reveal significant insights. In one study, we observed that Corn Variety 2 outperformed Corn Variety 1 during the final growth stage, leading to higher yields due to increased sugar and carbohydrate accumulation.
Figure 6: NDVI varying over time
In another example, NDMI values for two plants demonstrated similar trends, but Plant 1 experienced a notable drop during a drought period. This analysis underscores the importance of time-series data in evaluating plant resilience.
Our proprietary Botanic Analyzer model leverages these time-series techniques to provide detailed plantation assessments.
Figure 7: Botanic Analyzer
The model is trained based on both public data and private data sourced by the SC team. The labelled datasets along with the features constructed from the satellite data are used to produce the highly advanced Botanic Analyser model based on transformer networks.
We use transformer networks for the Botanic Analyser due to their exceptional ability to handle complex, high-dimensional data and effectively analyze time-series information. Transformers excel in capturing long-range dependencies within sequential data, making them ideal for processing the temporal variations inherent in satellite data. This capability allows the Botanic Analyser to accurately track changes in vegetation health, growth patterns, and environmental impacts over time.
Moreover, transformers are adept at integrating multi-band data, which includes various spectral bands captured by satellites. By leveraging attention mechanisms, the Botanic Analyzer can focus on the most relevant features across these bands, enhancing its ability to discern subtle differences in vegetation types and conditions. This multi-band data integration is crucial for producing precise and comprehensive analyses of botanical environments, ultimately leading to more informed and effective decision-making in agriculture, forestry, and environmental conservation.
Large-Scale Vegetation Mapping and Analysis
SC Solutions has executed extensive vegetation mapping and analysis projects on a national scale, including significant work near Riyadh, Saudi Arabia. These projects have demonstrated that certain plants exhibit superior resilience and growth characteristics compared to others.
Figure 8: Plant health under the same climate condition
Join Us in Our Mission
Are you interested in learning more about our innovative approaches and success stories? Contact the SC Solutions team to explore how we can collaborate to combat climate change through advanced remote sensing technologies.
