Friday, August 31, 2018

Tropical Rainfall Measuring Mission (TRMM/NASA) rainfall forecast for 27 June for South Asia

Rainfall Update: Rainfall declining in coming days (June 24-27, 2013)

Please see the visualization of rainfall intensities in the form of maps in the following pages.  The colored patches are the areas where heavy rainfall occurred in the past and is expected in the coming week. The red colored areas are where minimum rains and the purple ones are the areas where maximum rains occurred. You will see two kinds of maps (a) forecast for June 27, and (b) trend of rainfall.

1. Forecasts

Forecast for June 27, 2013 (72 hrs from June 24, 2013, 0600 UTC)

This is rainfall prediction for June 27, UTC 1200. Not so heavy rains expected in the region. The red colored areas represent average 100mm and yellow colored dots represent 200mm rainfall for  May 27, 2013 at 1200 UTC.

2. Recent rainfall trend

24 hours accumulated rainfall on June 24 by UTC 0600

Minimum 35 mm, central region of Nepal,  India, Bangladesh, Thailand and Cambodia and has almost no variability of rain.
3 days  accumulated rainfall on June 24 by UTC 0600
Minimum 100 mm, max 200mm in Brahmaputra region of Tibet, Bangladesh and Cambodia.

7 days  accumulated rainfall on June 24 by UTC 300

Patches of minimum 200 mm, max 350 mm  over Brahmaputra region, Myanmar and Cambodia. Rains are in declining trend.

3. Conclusion on Rains: 
The one day, three days and 7 days accumulation on June 24, at 0600 UTC trend shows rains are in declining order in the region. The forecast for June 27 shows regular monsoon rainfall trend.

4. Potential Flood and Landslide areas

4.1. Potential floods 

Expected in Bangladesh, Brahmaputra and Cambodia.

4.2 Potential landslides on June 24, 2013

Landslides very likely in the Brahmaputra region in Tibet.

4.3 Potential floods and landslides due to three days accumulation

Landslides just likely in South-east Tibet.

4.4 Potential floods and landslides due to 7 days  accumulation

Likelihood of landslides only in the south-east Tibet area.

5. Conclusion on Floods and Landslides:

Although both flood and landslide potentials still linger over Nepal and mainly in other parts in the region, they are in declining trends due mainly to the declining rainfall patterns for 72 hours.
For more on TRMM, please visit: To understand the process, please refer to this literature: Algorithm 3B42 - TRMM Merged HQ/Infrared Precipitation;
Note: These interpretations are based on some of NASA/TRMM experimental products. These experimental products usually imply their predictions may vary by a great margin of error, or may often predict very accurately. So, please exercise caution. These should be considered as educational products and the discussion should be taken in the same line. But these material have loads of scientific information to contemplate on.

June 24, 2013

Draft Remote Sensing Satellite Views of the Sun Koshi Landslide

Remote Sensing Satellite Data Visual Analysis of the Sun Koshi Landslide 

Indra Sharan KC

National Remote Sensing Center of Indian Space Research Organization, Government of India published Satellite Images of recent Landslide on Sun Koshi River in Nepal.

Huge peace of mountain slope The landslide had taken place on August 2, 2014. Nepal Army had been using light explosives, heavy equipment and muscle power to clear open channel to release water from the 2.36 km long, 200 m wide lake with a surface area of 44 Ha formed from the 5.5 million cubic meter debris deposit on Sunkoshi's bend in Jure. The landslide killed 156 people and destroyed 33 houses; a 2.5 KW hydropower plant in Sunkoshi Bazar; intake structure of 10KW Sunkoshi hydro in Lamosangu; transmission lines and towers and about 1.5 KM of road in Ramche and Jure. There was 67 MW of power less in the grid, due to this landslide which added extra 4 hours of load shedding in the regular load shedding of daily 10 hours. Significant amount of agricultural land and forest was impacted by the mega landslide.

The Indian Remote Sensing (IRS) Satellite images provide a great deal of valuable visual information. First, there was no areal or space view to inform scientists and concerned people on the scale and magnitude of the landslide and its impact area. Second, the data on landslide and lake parameters were all speculative.  The images answer these two questions very precisely.

There are two types of  IRS Images provided by the NRSC on Sun Koshi Landslide.

LISS IV - multi-spectral images ( Spatial resolution: 5.8 meter at nadir)

A LISS-IV sensor operates in multi-spectral and Mono mode. This  sensor provides data with a spatial resolution of  5.8m (at nadir) in both the modes.
In the Mono mode, the sensor provides data of a single band in the visible region, with a swath of 70 Km (at nadir) and across track steerability of + 26
degrees. In the multi-spectral mode, the sensor provides data of three bands with a swath of 23Km.

When the lake was fully formed:

 were taken by RESOURCESAT on 5th August. Here is the link to the images. The measurements data on the landslide and lake are given on the image itself. The Indian National remote Sensing Center LISS IV images are here.

The landslide width at the top and bottom portions measured 506 and 853 meters.  The width of the dam (across the river) measures to be 200 meter.
Similarly, the  length of the lake (straight from the dam to to Sunkoshi bazar at the confluence of Sun Koshi and Bhote Koshi rivers) is shown to be 2.36 KM. From the measurement on image, the area of the lake is 44 Ha.

The measurement is done on-screen using geo-referenced satellite image, and the value can change slightly from person to person. Perhaps these measured values are close to themselves 95% of the time.

Post- dam opening (post Sept 7) image:

The landslide lake suddenly eroded the channel created by the Army and 60 percent of water released in early hours of September 7, 2014. Not serious damage was noticed, although the villages just downstream were evacuated by the people themselves for fear of being washed away by the landslide lake outburst flood. The image clearly shows the decrease in water level. The upper portion of lake (the area near Sun Koshi bazaar, the confluence of Sun Koshi and Bhote Koshi)) is now out of water. Similarly  the area near the dam where a small stream by name Dabi Khola meets Sun Koshi, is also out of water impoundment. Village called Tekan, at this confluence, which is about a kilometer away from the landslide, was destroyed by the landslide and formation of the lake.  This image of September 17, clearly shows this change.

CartoSat Panchromatic Images: (Spatial resolution: 2.5 meter at nadir)

National Remote Sensing Center, Hyderabad also acquired CartoSat-2 data over the landslide on August 23, 2014, when the lake was still intact. Cartosat data are high resolution (2.5 meter), and the measurements on this data yields better results.  This image provides a great view of the landslide and the extent of the lake.  Satellite CartoSat carries two PAN sensors with 2.5m resolution and fore-aft stereo capability. The payload is designed to cater to applications in cartography, and terrain modeling.  However, the NRSC has only provided measurement data using LISS IV MX data, which is 5.8 meter in resolution.  

This image although has some clouds at the upper part of the landslide, has best captured the landslide and the lake in great details.

Quantitative measurements of Sun Koshi Landslide, Dam and Lake 

The high resolution image data is made available in JPG format. These are already processed graphics files. However, it was possible to geo-reference the image files using coordinates from topographic maps and the google earth satellite image data. A second order polynomial transformation was applied to the image after taking about 10 identifiable points on both reference and the source image. Then, the image is projected into a transverse mercator projection creating a metric image in which measurements can be carried out.

The data from the satellite imagery provides the following information:

Length of Landslide:  

The length of the longest portion of landslide on plan view is 1.5 km. NRSC using LISS IV image  labels it 1.3 KM. 

Width of the landslide:

The width at top is about 500 meter. The width at the bottom part is 900 meter. The landslide tapers from 900 meter at the bottom to 428 meter at the bottle neck portion above at 1.14 KM from the river bed. The average width of the landslide is 600 meter. The NRSC's LISS IV data is labelled as: 506 and 853 meters.

Gradient of the landslide:

The topographic elevation of the head of the landslide: 1500 masl
Elevation of bottom part of the landslide (at the Sun Koshi river course): 850 masl
Vertical height of the landslide: 1500-850 = 650 meter
Average slope of the landslide: 650 / 1500 = 1 meter vertical, 2.3 meter horizontal.

Landslide Dam:
Height (most of the reports) : 55 meter
Dam length: 300 meter
Length of toe of the dam: 700 meter
Gradient of the toe: 1:6 (calculated from photographs taken in the field).
Width of dam: 150 meter.

Landslide Lake:

Straight line distance as given by LISS IV data: 2.36 KM.
Length along the lake: 2.5 KM (calculated from the CartoSAT Data)
Perimeter: 6520 meter.
Area: 52 Ha (LISS IV as reported by NRSC: 44 Ha).

Indra Sharan KC