Effect of Mobile Phone Utilization in Predicting Drought Progression in Arid and Semi-Arid Lands: A Case of Kinna Ward in Isiolo County, Kenya

Abstract

In the event of natural disaster occurrences such as hurricanes, earthquakes, floods, pandemic and epidemic diseases, phone utilization aspects such as calls, SMS, airtime, and mobile money transaction have been widely used to monitor behavioral change and prevalence trend of the affected population. However, little is known on the ability of the meta data produced by network operators if could be utilized to monitor and predict drought progression in arid and semi-arid areas in Kenya. The gap necessitated the conveyance of this evaluation to give more insight on the possibility of utilizing the phone generated records in monitoring drought progression. The study was a panel data conducted on January, 2020, June, 2020 and November, 2020 on 233 phone users sampled from a population of 4345 phone owners in Kinna ward, Isiolo County, Kenya. The three drought phases: normal, emergency, and recovery were developed and projected based on the area standard precipitation index derived from 20092019 precipitation values recorded by the Kenya meteorological department. Daily information on calls made and received, SMSs send and received, airtime purchased, and money send and received through mobile money as well as villages visited during the three phases were recorded accordingly. The study used fixed effects panel regression model to reveal the fixed effect of phone utilization aspects on drought progression phases and found out that the phone utilization aspects: received calls, call length, money transaction records and locations significantly influenced drought progression from normal, to emergency to recovery. The study concludes that telecommunication data which can be cheaply obtained has a potential to be employed in drought monitoring in regions with restricted ability to gather information like in ASALs regions of Kenya. Keywordsdrought phases, phone utilization aspects, disaster, data 1.0 INTRODUCTION Kenya’s arid and semi-arid lands (ASAL) covers over 75% of her land, with 29 counties out of 47 categorized under ASAL region. The ASAL face unique challenges that affect peoples’ socioeconomic development and environmental sustainability [1]. These challenges include climate change, insecurity, communal conflicts, low capital investment, poverty prevalence, and food insecurity which require a multidisciplinary approach in search of sustainable solutions. The ASAL is characterized by low and erratic precipitation and sporadic drought occurrences. The droughts can vary in severity, but the region is no stranger to these catastrophes. Between 1900 and 2010, more than 18 severe drought periods were experienced in the region s history [1]. According to Stockholm Environment Institute (2009) report on the economies climate change Kenya, the financial loss of the 1998-2000 drought that affected Kenya was documented at $2.8 billion. More significantly, the assessment undertaken after the 2008-2011 drought estimated that $12.1 billion were lost in countries national economy. Such huge losses were evidenced due to cases of local market distortions, reliance on relief aid, the dilemma of severe drought followed by heavy rains that causes floods in the regions and destruction of resourceful infrastructure [2]. The drought and flooding tragedies occurrence further distorted the economic activities as it was experienced in 2018 drought and floods which destroyed road infrastructure, food supplies for the residents due to flooded road network making them impassable. While humanitarian assistance can save lives, long-term strategies should be put in place to build communities resilience in order to manage drought crisis in real time and help them prevent future destruction [2]. According to Kenya’s national 2019 census statistics, approximately 20 million individuals own a mobile phone. Whereas 22.6% of persons aged above 3 years have access to internet [3]. During the pre-event(normal) phase of drought classification, the situation is characterized by community preparedness actions where cell phones and other media networking instruments like radio and TV can give data about the changing behavior in terms of phone utilization variation and other communication information toward risk and tips to promptly plan for its effect. During the reaction(emergency) and recuperation(recovery) stages, the general media always considered on giving evaluation information of damages and misfortunes caused by these climatic disasters. Because of the farreaching acknowledgment of the expanding multifaceted nature with respect to disaster mitigation, there has been enthusiasm for developing resilience plans for drought monitoring for ASAL. This calls for extended financing from relevant actors and the incorporation of proactive drought management as a center planning component across all sectors. The ASAL region therefore remain a sole proof for developing enthusiasm in drought monitoring and management [4]. While the underlying diversion in drought assessment tools is a cause of concern, there is a growing anxiety to achieve an effective drought monitoring International Journal of Scientific and Research Publications, Volume 11, Issue 5, May 2021 726 ISSN 2250-3153 This publication is licensed under Creative Commons Attribution CC BY. http://dx.doi.org/10.29322/IJSRP.11.05.2021.p11376 www.ijsrp.org approach that can resolve all ASALs climate changes effects due to its catastrophe recurrent happenings. As resilience centered investments keep on being made in ASALs, the requirement for better understanding and monitoring continue to increase. In this manner, the pointers for drought progression and how dynamic would that prediction be, is the main concern to government agencies, humanitarian organizations, and environmental researchers [4]. Detailed datasets are available due to the evolution of the transformative and innovative telecommunication transmission industry. Cellphones have reformed the world of correspondence. Mobile phones have picked up a great deal of significance in the lifestyles of most Kenyans. In the region, mobile phone penetration in Kenya stands at 62 percent which is the highest [5]. The penetration has been enhanced by the possibility of individual customers to own multiple SIM cards an attribute motivated by subscribers wanting to gain benefits from low prices offered by competitor telecommunication companies [5]. Through various mobile-based applications, numerous individuals lives have been transformed [6]. Individuals currently get information, learning and conducting business using their mobile phones. Additionally, there are numerous advancements that are being done on the broad applications that identifies call detail records (CDR) as an indicator could offer one of the best avenues for effective and efficient drought monitoring and management, particularly with the increased mobile utilization in Kenya’s disaster-stricken counties [5,6]. In their study [7], revealed that regular drought early warning further guarantees household food safety and nourishment by availing timely data, however, the provision of these early warnings faced the high cost and irregular constrains. Their study recommended an improvement of existing early warning systems which focused primarily on more aggregate indicators mostly around Household Economic Approach (HEAs), which CDRs could have provided as well. An information gap that necessitated the survey. 2.0 EMPIRICAL OVERVIEW 2.1 Mobile Data Utilization in Disaster Management Numerous studies have dealt with the era of big data application in diminishing tragedies. Relevant literatures are available which have successfully attested use of mobile generated records as an indicator for social interactions [8]. Telecommunication networks extensively collects subscribers’ data on calls, calling trend, financial transactions and their location [2,8]. The volume data derived from mobile users can be abused for erroneous intention. However, scientists have an enormous opportunity to utilize the privacy-safe anonymized datasets to discover the configurations and dynamics of mobile users at diverse stages of communal unique behavior with precision [2]. Total elimination of disaster event occurrences is beyond human capability, however, innovative technologies are accredited as approaches that can reduce to some extent the magnitude of tragedy losses. Emergency responders often hears about the crisis occurrences through bystanders’ accounts either on TV coverage or callers of the emergency numbers. They provide information which lacks wide range perspective of the actual status of the crisis and therefore might not be reliable and lacks the aspect of actionable data [9]. Managing catastrophe events require actual facts dissemination being utilized by emergency service providers in order to target the real victims at the accurate time. The concern is being focused on available critical information and innovation to avert, alleviate, and manage calamity [8]. Due to technological advancement, real-time data can be captured in various sources such as social networks and mobile devices [2]. Situational analysis using the advance technologies in big data analytics gives accurate information that empower emergency service providers in making informed decision, taking suitable act as well as improved management and response process [8]. The Covid-19 pandemic outbreak has intensified the discussion on usage of mobile phone generated records in epidemic response [10]. The mobile data generated is still being utilized to curb corona virus blowout through informing the response team on human mobility variation, evaluation of intervention as well as identifying hotspots where innovative actions need to be applied [10]. 3.0 RESEARCH ELABORATIONS 3.1 Study Design The survey explored the possibility that variability of phone generated records: calls, SMS, airtime, mobile money transaction and location can b