September/23 Report
Last updated
Last updated
The study area comprises the "Fazenda Floresta Amazônica" located in the municipality of Apuí - AM (Figure 1). According to the IBGE vegetation classification (2021), the area of the farm falls into two types of vegetation: Open Ombrophilous Forest (OOF) and Dense Ombrophilous Forest (DOF).
This technical report uses NDVI to monitor possible changes in vegetation within the area known as "Fazenda Floresta Amazônica", as well as detecting possible degraded and deforested areas. In addition, using images of deforestation alerts from the INPE and Imazon database, deforestation alerts were monitored for the period from September 1 to 30, 2023, both in the area of the farm and in its surroundings. In remote sensing, monitoring changes in vegetation cover at regional and global scales, the normalized difference vegetation index (NDVI) is commonly used (Shi et al., 2021, Wu et al., 2020). NDVI is an important indicator that reflects the state of vegetation, which is affected by precipitation, anthropogenic activities, temperature and soil water content.
In order to draw up the NDVI map for the period from 01/09 to 15/09 and 16/09 to 30/09, there were no ideal atmospheric conditions for images from the Amazonia 1 satellite. We therefore had to resort to other satellites, and only the Landsat 9 satellite image from 02/09 showed conditions for drawing up the NDVI map, with 0% cloud cover and the image from 26/09 with 10% cloud cover.
Following the usual procedure, the NDVI was obtained by applying the algorithm which is based on the difference in near infrared reflectance and red reflectance divided by the sum of the spectral bands (1), where NIR is the infrared electromagnetic spectrum and R is the red electromagnetic spectrum, varying between -1 and 1 (ROUSE et al., 1974), as shown in the equation: NDVI= ((NIR-R) / (NIR+R))
Landsat 9
02/09
2023
B4-B5
Landsat 9
26/09
2023
B4-B5
Table 1. Specifications of the satellite image used to calculate NDVI.
In addition, a SAVI map was drawn up to assess and prove that there is no deforestation in the study area. The Soil Adjusted Vegetation Index (SAVI) is a variation of the Normalized Difference Vegetation Index (NDVI) that considers the influence of the soil on the spectral response of plants. It is used to estimate the density and vigor of vegetation by correcting for the effects of the soil present in the satellite image.
Where: NIR is the reflectance value in the near infrared band; RED is the reflectance value in the red band and L is the soil adjustment factor, which varies from -1 to 1 (usually set at 0.5).
Including the soil adjustment factor (L) in the formula makes it possible to compensate for the influence of soil on reflectance, especially in areas with sparse vegetation or exposed soil. The value of L is generally set at 0.5, but can vary depending on the characteristics of the vegetation and soil in the study region.
In order to gain a better understanding of the indices used in the study area, the Normalized Burn Ratio (NBR) map was also drawn up in Qgis. This is an index widely used to detect areas affected by fires (Key and Benson, 1999), deforestation or damaged vegetation. It uses the near-infrared (NIR) and mid-infrared (MIR) bands of satellite images to map the extent and severity of fires. This was done using Landsat 8 satellite images of bands 5 and 7 from 26/09/2023 at 14:20 hours (https://earthexplorer.usgs.gov/).
Where: NIR is the reflectance value in the near-infrared band and MIR is the reflectance value in the mid-infrared band.
The Normalized Burn Ratio (NBR) varies on a scale of values according to the methodology and scale used. However, a common convention is for the NBR to have values from -1 to 1. Negative NBR values (-1 to 0) generally indicate areas affected by recent fires, deforestation or damaged vegetation. Values closer to -1 indicate greater severity of burning or damage to vegetation. Positive NBR values (0 to 1) are often associated with areas with healthy vegetation. Values closer to 1 indicate greater vegetation vigor
To assess the advance of deforestation around the "Fazenda Floresta Amazônica" area, a radius of 50 km was determined. Deforestation alert databases from INPE and Imazon were used. The data obtained for the month of February was in shapefile format and all the processing and filtering of the alert points within the radius was carried out using the Qgis software.
Imazon's Deforestation Alert System (SAD) currently uses NASA's Landsat 7 and 8 satellites and the European Space Agency's Sentinel 1A, 1B, 2A and 2B satellites. SAD detects forest degradation or deforestation that has occurred in areas of 1 hectare or more.
INPE uses images from the WFI sensors of the Sino-Brazilian Earth Resources Satellite (CBERS4) and AWiFS from the Indian Remote Sensing Satellite (IRS), with 64 and 56 meters of spatial resolution respectively.
The data is sent daily to the Brazilian Institute for the Environment and Renewable Natural Resources (IBAMA) with no restriction on the minimum area mapped. However, the polygons are made available to the general public with a minimum size of 6.25 hectares, thus allowing a comparison criterion to be established with the data generated by the PRODES project.
The pattern of forest cover change is identified by visual interpretation based on five main elements (color, tone, texture, shape and context) and uses the Linear Spectral Mixture Model (LSMM) technique, together with its multispectral image in color composition to map the following classes:
Deforestation: deforestation with exposed soil and deforestation with vegetation and mining; Degradation: degradation, forest fire scar; logging: selective logging type 1 (disordered), selective logging type 2 (Geometric).
The spatio-temporal distribution of the normalized difference vegetation index (NDVI), calculated using Landsat 9 images from September 2nd, showed that the values ranged from -0.09 to 0.5. The lowest NDVI values, between -0.09 and 0.3, correspond to the rivers and small clearings formed by the rivers. Values between 0.4 and 0.92 correspond to forest areas (Figure 2). The NDVI values for September 26 ranged from -0.01 to 0.55 (Figure 3).
Comparing the NDVI maps for September with the NDVI map for August, the values for both months were not similar. The NDVI values in September were different from those observed in August. However, this variation is mainly attributable to the vegetation's natural adaptation to the dry period characteristic of this region.
An important observation made using NDVI during the dry season is related to the change in leaf color. Some trees may lose part of their leaves, which turn yellowish or brown. This change in color influences NDVI values, which tend to decrease as the leaves lose their ability to absorb light in the spectral range relevant to NDVI. In any case, the NDVI calculated for this area indicates that the forest is extremely homogeneous and has no degraded or deforested areas.
In the context of the Amazon, monitoring NDVI values is fundamental for assessing the impact of human activities, such as deforestation, on the health of the forest.
NDVI can also be used to identify areas with greater potential for vegetation conservation and recovery, contributing to the planning of more effective public policies for forest conservation.
In the Soil Adjusted Vegetation Index (SAVI), the value ranged from -0.02 to 0.82. These results reinforce the conclusion that there was no evidence of deforestation or any human activity in the study area. The value of -0.17 to 0.3 corresponds to bodies of water (rivers) and areas exposed to riverbanks, due to the river level being naturally lower at this time of year. While values greater than 0.4 correspond to areas with dense forests in a good state of preservation.
Regarding the normalized burning ratio, the values on the map ranged from -0.02 to 0.55, indicating that the study area shows no signs of burning. Values close to zero and/or negative are represented by rivers (bodies of water) and clearings, such as sandbanks, while values above 0.3 are represented by dense vegetation.
When comparing the results of the normalized burning ratio with the previous month, it was observed that the values are similar, indicating that there has been no deforestation or burning within the study area.
Regarding to deforestation alerts, INPE and Imazon found a total of 6 alert points within a 50 km radius. The satellite that detected the largest number of points (Table 2) was CBERS-4. The areas covered by the alert points ranged from 6.32 to 17.58 hectares, which indicates that deforestation is occurring on a considerable scale.
1
13/09/2023
CBERS-4
766478
9049119
11,7
2
13/09/2023
CBERS-4
765928
9049580
15,2
3
16/09/2023
CBERS-4
777672
9159303
11,74
4
23/09/2023
CBERS-4
777774
9048449
6,32
5
24/09/2023
Amazonia-1
774803
9156778
17,58
6
28/09/2023
CBERS-4
772352
9043113
6,8
There is no evidence of deforestation and/or forest degradation inside the farm from the NDVI, SAVI and NDR maps in September
From the INPE and Imazon deforestation alerts, it was possible to identify 6 deforestation hotspots in September.
July and August were the months with the lowest number of deforestation alerts.
It was possible to identify that there was no increase in the deforested area of the alert point closest to the farm (3 km – Paxiúba River).
Cunha, A.P.M., Alvala, R.C., Nobre, C.A., Carvalho, M.A., 2015. Monitoring vegetative drought dynamics in the Brazilian semiarid region. Agric. For. Meteorol. 214, 494–505. https://doi.org/10.1016/j.agrformet.2015.09.010.Chu, H.S., Venevsky, S., Wu, C., Wang, M.H., 2019. NDVIbased vegetation dynamics and its response to climate changes at Amur-Heilongjiang River Basin from 1982 to 2015. Sci. Total Environ. 650, 2051–2062. https://doi.org/10.1016/j.scitotenv.2018.09.115.Valle Júnior, R.F., Siqueira, H.E., Valera, C.A., Oliveira, C.F., Fernandes, L.F.S., Moura, J. P., Pacheco, F.A.L., 2019. Diagnosis of degraded pastures using an improved NDVI-based remote sensing approach: an application to the environmental protection area of uberaba river basin (minas gerais, Brazil). Remote Sensing Applications: Society and Environment 14, 2033. https://doi.org/10.1016/j.rsase.2019.02.001.Xu, H.J., Wang, X.P., Yang, T.B., 2017. Trend shifts in satellitederived vegetation growth in Central Eurasia, 1982–2013. Sci. Total Environ. 579, 1658–1674. https://doi.org/10.1016/j.scitotenv.2016.11.182.Botucatu (SP), 12 de outubro de 2023.(11) 94004-0092 alessandro@zabottoambiental.com.br www.zabottoambiental.com.br
