Sunday, September 14, 2014

Landslide detection using AutoMCU on Landsat images

I recently took an online course on CLASlite offered by the Carnegie Institution for Science.  CLASlite is a software designed to automatically detect deforestation, forest degradation and forest regrowth. It uses the AutoMCU algorithm to detect the percentage of various types of vegetation, allowing to automatically detect deforestation using Landsat images.

In one of my blog posts, I used Landsat image to see the Rockfall scar on Annapurna mountain. I detected the scar manually, but imagine if we could do this automatically. Manually delineating landslide takes a lot of time. We need a lot of historical landslide data to build better landslide hazard maps and landslide prediction models. Being able to automatically detect landslide would be very useful to the geo-disaster community.

A useful index used in remote sensing is the Normalized Difference Vegetation Index (NDVI). This is used to detect vegetations. Most of the hilly region in the Himalayan region is covered with vegetation. When a landslide occurs in such a vegetated area, the vegetation will be swept away, leaving fresh bare soil. Thus, could a change in NDVI be used as an indication of landslide?

Not necessarily. Deforestation would also have a similar effect in NDVI, thus it would be difficult to differentiate landslide from deforestation or another normal bare surface. However, there is a peculiar difference between a fresh landslide and existing normal bare surface. Fresh landslide generally consists of no vegetation, but an existing bare surface would contain some non-photosynthetic vegetations like chopped woods, dead leaves or other dry plans. This is when I thought AutoMCU can come in hand.

Unlike NDVI, AutoMCU outputs three levels of indicators: Bare Substrate (BS), Photosynthetic Vegetation (PV) and Non-photosynthetic Vegetation (NPV). NDVI can only detect photosynthetic vegetation, however AutoMCU can distinguish Bare Substrate from Non-photosynthetic Vegetation. I wanted to try this out myself, so I used the Landsat 8 images near the recent Jure landslide site in Sindhupalchowk, Nepal.

Saturday, July 19, 2014

Intrinsic Anisotropic Poisson Effect that exists in the basic Applied Element Formulation

In my previous article, I had discussed about the AEM program for linear 2D materials that I had written in Python. Although, poisson's ratio was not considered in that program, I found that when elements are compressed (or stretched) diagonally, an intrinsic poisson effect is induced. However, such poisson effect is non existent when elements are compressed (or stretched) orthogonally. This is mainly because when the elements are compressed diagonally, the diagonal displacement induces extra force on the two lateral elements as shown in figure 1.

Note: Element properties used in all of the simulations presented here are: 
Young's Modulus (E) = 207 GPa
Shear Modulus (G) = 79.1 GPa
Element Dimension (a by b) = 0.1 m by 0.1 m
Element Thickness (T) = 0.15 m
Load (L) = 1000 kN [each red arrows]
Also in all of the figures presented here, green squares represent undeformed elements and blue squares represent deformed elements. Displacements are magnified 500 times.
Figure 1: Lateral displacements induced (represented by black arrows) when force (represented by red arrows) is applied in a direction diagonal to the element edges. This is the intrinsic anisotropic poisson effect induced in the basic applied element formulation. 

Tuesday, March 4, 2014

Formulating the Applied Element Method: Linear 2D (Part I)

I first came to know about the Applied Element Method (AEM) during the one day seminar on "Geotechnics and Geo hazards" 2012 held in Nepal. Ramesh Guragain had mentioned this method in his presentation on the study of collapse of masonry structures (and something to do with fragility functions I believe). The results he showed using AEM were really compelling. I wanted to learn more about this method.

Here I discuss the basic procedures that I have followed to develop an AEM program on Python that can solve Linear 2D structural problems. Poisson's ratio is not considered here. All of the procedures discussed below are based on Meguro and Tagel-Den (2000). The developed program is then used to solve a classic cantilever beam problem. The displacement values obtained using AEM are compared with corresponding theoretical values and values obtained from Finite Element Method (FEM) for plane stress condition. AEM performs better compared to FEM even when the number of elements is small.


AEM is modeled by dividing the structure into rigid elements connected with pairs of normal and shear springs as shown in figure 1. The stiffness values of each pair of springs represent the material property of certain area of both the elements as shown in figure 1 b. The stiffness values are determined as shown below:

where E and G are Young's and shear modulus, d is the distance between springs, T is the thickness of the element and a is the length of representative area.

Friday, October 18, 2013

Simple 2D simulation of the Phailin Cyclone

It was the time of Dashain festival here at Nepal in 2013, which falls during the autumn season. The rainy season was supposed to have ended by then, but it had been raining for a couple of days and it was mainly because of the Tropical Cyclone Phailin that had hit the south eastern coast of India. No, the cyclone didn't reach Nepal, but it did send clouds here.
Source: Cooperative Institute for Meteorological Satellite Studies

Here is my simplified 2D simulation of the Phailin Cyclone I did using the instantaneous vorticity data as initial condition. I mainly solve the 2D Navier-Stokes equation using a forth order finite difference scheme in MATLAB.

Monday, July 8, 2013

A scar on Annapurna that had caused the Seti Flood of 2012

I had been going through different articles and papers about the Seti Flash Flood of 2012 and, I wanted to try out something for myself.

On May 5th of 2012, a flash flood full of debris, triggered in the Seti river, had devastated different parts of Pokhara with an estimated 71 people killed. A unique thing about this flood was that, it was probably triggered by an avalanche near the Annapurna range. This triggering of the avalanche was actually captured in tape by a Russian pilot who was flying over the same area at the time of avalanche, and about 40 minutes after the avalanche, the devastating  flood hit Pokhara. One of the most comprehensive reports on the cause of the flood should be the report prepared by Dr. NP Bhandary, Dr. RK Dahal and Prof. M Okamura based on their field visit (Bhandary et al. 2012).

Initial speculations on what might had caused the flood ranged from intensive rainfall-cloudburst flood, possible GLOF (Glacial lake outburst flood) as well as LDOF (Landslide Dam Outburst Flood). But no significant evidences supporting the latter two were found in the field. Also, the flood had occurred way before the monsoon, reducing the possibility of cloudburst flood . Looking at the devastation and magnitude of the flood, initiation of such a flood would require a huge source of water (millions of cubic meters) and these speculations couldn't properly explain the source of such a huge quantity of water.