OUR INSPIRATION:

Peru is a country heavily affected by El Nino. El Nino is a climate phenomena that involves the surface of the ocean warming up and a change in atmospheric circulation (3), leading to heavy rainfall. A natural consequence of the rainfall are events we call “Huaicos”. Huaicos are essentially strong flowing mudslides. Through the years, Huaicos in Peru has led to the loss of many homes, loss of lives, but most importantly, loss of essential facilities and services. This challenge gives an opportunity for us to create a solution to minimize and prevent the level of destruction and economical loss that the situations such as the “huaicos” create.

THE APPROACH:

The first step towards achieving the objective of the challenge was obtaining database sources. Essentially, using the database provided by satellites COPERNICUS SENTINEL-1A and ALOS-2, we were provided with a collection of imagery from different parts of the world. COPERNICUS SENTINEL- 1A makes available images taken by a satellite resolution of 5x5m.This is enough resolution to engage the system of recognition imagery operated by Google Earth that can be used to identify the critical infrastructure present in a localized point. With this open for use, we can proceed to develop an algorithm that would implement each different recognized critical facility with an assigned monetary value (The value index for each critical infrastructure is a data set provided by each state’s own government due to legislative regulations and access restriction).

Next, our attention turns to the quantification of damage. To define a parameter that details the range of damage caused by a flooding, data sets given by ALOS-2 will be used. This data set would then be processed through a system provided by the Jet Propulsion Laboratory, ARIA-SG, whilst using InSAR Scientific Computing Environment as the processor for imagery data (4) [This process of imaging processing with SAR was taken directly from a study done wit ALOS-2. References are below.]

With such, it would then become possible to quantify the percentage of damage caused towards critical infrastructure after an occurrence. Thus, it becomes now possible to relate the meteorological variables with the amount of structural damage caused in a critical infrastructure. 

Meteorological variables are themselves defined by the use of IBM Weather Channel data sets. 

With a monetary variable and the meteorological parameters defined it is possible to correlate both through the aforementioned damage variable. Hence, meeting the conditions to create a predictive model that allows us to know economic losses through specific meteorological parameters.

A diagram is included below that summarizes the whole process:

OUR TOOLS:

• To achieve this we required a strong internet connection is required along with a compiler that supports python3 (Anaconda or Google Colab recommended).
• PC with a GPU that supports convolutional networks
• The ability to have constant connection to the GOOGLE EARTH API.
• Jupyter Notebook - For the execution of referencial algorithms (see later)

WE ARE PROUD:

Here in Piggybank, we are proud of what we managed to design, develop, and execute with the short amount of time given. Problems faced during the developing stages occurred consistently as the methodology by itself becomes really complex with the restriction of data access and the limitation of skill with coding.

For more information: http://piggybankcom.co 

Piggybank prioritized the use of satellite data and imagery collection. To achieve our goals, we chose to use data sets from two different satellites, ALOS-2 and COPERNICUS SENTINEL- 1A. ALOS-2 is a satellite provided by JAXA, whilst COPERNICUS SENTINEL-1A is an ESA covered satellite. Through our project ALOS-2 provides SAR imagery that we can use to create Flood Proxy and Damage Proxy Maps through the processing of the image through ARIA. COPERNICUS SENTINEL-1A, is an observational satellite that also includes SAR imagery, but what we mainly take from its functionality, besides data imagery, is the integrated recognition system it has. The data sets obtained by both of them influenced the overall trajectory of our project as we wanted to create a methodology that revolved around the open-accessed data provided. However, we discovered, through the development stages of the project, that the open-accessed data specifically available for this challenge was severely limited, hence the reason why we are proposing a methodology and not exposing a full working program.

Additionally, we also used NASA's Advanced Rapid Imaging and Analysis (ARIA), as the primary processing method for both satellite's data collection. We received inspiration to do this by a paper published by a group of scientists analysing a case study of SAR imagery through detection method after a typhoon in Japan. This can be found in ARIA's Publication Page.

These three data sources allowed us to reinforce our model through the construction of the rest of the parameters around the fundamental programmed basis these three programs used. Furthermore, we also used Google Earth's APIs to introduce the idea of a recognition system that would distinguish critical infrastructure and set them apart from non-critical structures.

Referenced code from Case Study of Typhoon Hagibis:

https://github.com/earthobservatory/slcp2pm.

ARIA: https://www.aria.jpl.nasa.gov/publications.html

ALOS-2: https://disasters.nasa.gov/programs/alos-2

COPERNICUS SENTINEL-1A:https://www.copernicus.eu/en

For more information: http://piggybankcom.co 

SLIDES: https://docs.google.com/presentation/d/1dyWDMb-Qc1KS5yRuOyX5OiXKD2UxRLKN92iB5u-4atQ/edit?usp=sharing

(1) “A Flood of Ideas.” 2020.Spaceappschallenge.org, 2020, 2020.spaceappschallenge.org/challenges/confront/flood-ideas/details. 

(2) UNDRR. (2018). Monitoreo de la aplicación del Marco de Sendai para la Reducción del Riesgo de Desastres 2015-2030: Una instantánea de la presentación de informes para 2018. ONU. https://www.undrr.org/es/publication/monitoreo-de-la-aplicacion-del-marco-de-sendai-para-la-reduccion-del-riesgo-de

(3) Qué es El Fenómeno El Niño | Sistemas de Información Clima y Agua - INTA. (s. f.). INSTITUTO NACIONAL DE TECNOLOGIA AGROPECUARIA. Recuperado 4 de octubre de 2020, de http://climayagua.inta.gob.ar/que_es_el_fenomeno_el_ni%C3%B1o

(4) WJ Tay Yun , Chin ,Alok Bhardwaj ,Jungkyo Jung, Emma M. Hill. (2020, 25 marzo). Mapeo rápido de inundaciones y daños utilizando un radar de apertura sintética en respuesta al tifón Hagibis, Japón. Nature. https://www.nature.com/articles/s41597-020-0443-5#Equ2

(5) CENEPRED. (2014). Manual Para la Evaluacion de Riesgos Originados por Inundaciones Fluviales. Direccion de Gestion de Procesos Subdireccion de Normas y Lineamientos. http://www.congreso.gob.pe/Docs/OCI/files/manual_para_la_evaluaci%C3%B3n_de_riesgos_originados_por_inundaciones_fluviales.pdf

(6) JAXA. (s. f.). JAXA para la Tierra. Recuperado 4 de octubre de 2020, de http://earth.jaxa.jp/en.html#

(7)NASA. (s. f.). Instruments & Models. NASA EARTH SCIENCE DISASTER PROGRAM. Recuperado 3 de octubre de 2020, de https://disasters.nasa.gov/instruments

(8)ESA, EMSA, EUMETSAT, ECMWF, EEA, MERCATOR OCEAN INTERNATIONAL, & FRONTEX. (s. f.). Europe's eyes on earth. COPERNICUS. Recuperado 4 de octubre de 2020, de https://www.copernicus.eu/en

(9)Shinichi, S., & JAXA. (s. f.). Satélite Avanzado de Observación Terrestre-2 «DAICHI-2» (ALOS-2). JAXA. Recuperado 1 de octubre de 2020, de http://global.jaxa.jp/projects/sat/alos2/