Where Tall Things Are - TLG

1.Introduction

The purpose of this project is to propose new ways to use the data received from the ICESat-2 satellite. In order to highlight the usefulness of the ICESat-2 data, a new product for monitoring the desert biome is suggested. This product is for hot deserts with arid climate and not arctic deserts that are covered with ice or permafrost. The new product, titled “ATL_TLG” , will be a level 3A product, meaning it will incorporate the data produced by the ATL03 and ATL09 products.The ATL_TLG product will contain the height of the ground, the height of sand dunes, canopy height of the desert vegetation, slopes and slope lengths.

2.Data products

Even though deserts seem a small part of the human world, they cover approximately 20% of the earth's surface and one sixth of the human population inhabits them. They have a rich biodiversity and a significant role to the environment. Many of the plants and animals benefit humans, such as camels and the plant date, an important food source in North Africa and the Middle East. Furthermore, the Saharan dust contains phosphorus and other beneficial minerals that travel through wind currents to the Amazon rainforest providing it with valuable nutrients. More than 22.000 tons are estimated to travel annually.

Therefore, the monitoring of the desert biome is significant for the quality of life of millions. The ground height data for these areas can offer a more accurate way of mapping the topography and it’s changes than the SRTM.

Due to the extremely dynamic nature of the sand dunes and their negative impact on atmospheric quality, the ground height data of that nature was decided to be represented separately. Dust from the desert can travel thousands of kilometers and stay gathered in the atmosphere for many days. Despite the advantages this can bring to the Amazon rainforest, the concentration of dust particles can lead to health issues (especially for vulnerable groups) and can also interfere with the plant’s evapotranspiration. ICESat-2 can monitor the sand dune elevation more accurately than ever before. This information can be used in combination with other resources to predict the dispersion of the dust particles in the atmosphere.The following image shows a track of the ICESat2 Satellite and the plotted height of the received photons along the track’s latitude.

Image 1. ICESat 2 - Height of the photons received on the Saharan desert. Depiction of a sand dune.

Desert vegetation is not as scarce as generally thought. The canopy height data, the slopes and slope lengths are important variables that can be used to estimate vegetation mass and soil degradation, for example through the Universal Soil Loss Equation. Desert agriculture is a big source of food for the local population. In arid climates where water is scarce, desertification is a real threat. The combined use of these data is estimated to be important for risk management applications of these kind. The canopy height is intended to be used supplementary with the ATL08 product. In the following image we can see the difference in canopy heights for an area in the borders of the Sahara desert. The first plot is from data received in March and the second in December. The difference in canopy is visible and is due to the seasonal change. This example is to simulate the idea behind the identification of desertification using ICESat2 data since vegetation density is lost in the process.

Image 2. ICESat2 - Canopy heights.

A short description of the products provided can be seen in the following table.

Table 1. Short description of the products provided

3. Algorithm

On the ATL03 product the received photons are classified into two categories: the signal event photons and the background noise photons. As is also stated in the ATL08 product’s Algorithm Theoretical Basis Document the photons that originate from arid regions and are tied with the vegetation canopy, might be classified as background noise, due to the low density of the vegetation. The problem here is intensified due to the more inhospitable and arid conditions that exist mainly on sand deserts. Therefore, a change to the algorithm is considered for future work.

In order to focus only on the desired regions a mask for the desert areas is required. This is proposed to be done with the use of Radar images from the RADARSAT satellite, in order to detect soil moisture, with combination of the spectral reflectance of the usual land desert materials and climate data.

The remaining photons will have to be classified whether they originate from ground, sand dunes. To determine ground or sand dunes we can use :

1. microwave remote sensing. In order to identify land differentiation, values of scattering coefficient will be used.
2. hyperspectral / multispectral remote sensing: estimation of the different classes can be provided through spectral signatures in order to classify them or image segmentation can be applied .

As far as the canopy is concerned the algorithm is borrowed from the ALT08 product.

Slopes will be calculated with a linear regression of the terrain photons every 100m .

Slope length is the ramp length between the terrain photons every 100m.

4. Future work

Right now there is a difficulty in getting revisits of the desirable tracks mainly on the lower latitudes of the deserts. Plus, for the first 2 years of the mission the plan is to offset the tracks in vegetation regions. In order to concentrate on the deserts a request must be submitted as a Target of Opportunity so as to have more revisits. Project’s future plans is to create a more accurate Digital Elevation Model and gridded maps of the desert biome. By improving deserts monitoring, we take a step forward towards Earth Observation as a tool to improve life.