Network of Automatic weather stations : Time …metnet.imd.gov.in/mausamdocs/16539_F.pdfSystem description and measurement schema The remote field station consists of a data logger,

MAUSAM, 65, 3 (July 2014), 393-406

551.502 : 551.508

Network of Automatic weather stations : Time division multiple access type

M. R. RANALKAR, M. K. GUPTA*, R. P. MISHRA, ANJIT ANJAN and S. KRISHNAIAH

India Meteorological Department, Pune, India

*Regional Meteorological Center, Guwahati, India

(Received 26 February 2013)

e mail : [email protected]

सार ‒ दर दर तक फैले िजन के्षत्र से पे्रक्षणा मू ू क आँकड़ ेप्रा त नहीं हो पात ेह उनकी भरपाई करने और अिधकत ृ यिक् तय द्वारा संचािलत सतह पे्रक्षणा मक संजाल का संवधर्न करने के िलए अप्रैल 2009 से फरवरी 2012 के दौरान भारत म 550 वचािलत मौसम कद्र (ए ड यू एस) का संजाल थािपत िकया गया। वाय तापमानु , सापेक्ष आद्रर्ता, वायमंडलीय ुदाब, वषार्, पवन और भमंडलीय सौर िविकरण जैसे प्राचल के आकँड़ ेलेने के िलए सभी ए ड यू ू एस, संवेदक से लैस ह। मदा तापमानृ , मदा नमीृ , प त ेका तापमान और प त ेकी आद्रर्ता लेने के िलए 127 ए ड यू एस म अितिरक् त संवेदक लगाए गए ह। घंटेवार मापन के िलए डाटा लॉगसर् का प्रोग्रािमत िकया गया है। प्रभावशाली चैनल उपयोग के िलए 93.5° पवर् म थाू िपत इंसैट - 3ए भू- थैितक उपग्रह म डाटा िरले ट्रांसपॉ डर (DRT) के मा यम से डाटा के प्रसारण के िलए टाइम िडवीजन म टीपल एक् सेस (TDMA) नामक उपग्रह टैलीमीट्री तकनीक का प्रयोग िकया गया है। पणे म थाु िपत कद्रीय आँकड़ा संग्रहण भू-कद्र म डाटा प्रा त और अिभलेिखत िकया जाता है। भकद्रू म मा य समद्र सतह ताु पमान, ओसांक तापमान, दाब प्रवि तृ , दैिनक अिधकतम और यूनतम तापमान यु प न िकए जात े है। ड यू एम ओ सचना ूप्रणाली (WIS) तथा www.imdaws.com के मा यम से WMO कोड के प म अंितम उपयोगकतार्ओं के िलए लगभग वा तिवक समय म डाटा उपल ध कराए जात ेह। इस शोध पत्र म ए ड यू एस के तकनीकी यौरे, संजाल योजना, टेलीिमट्री प्रणाली के लक्षण, भकद्र आकँड़ा संग्रहण की क्षमताएं और संजाल के कायर्िन पाू दन के आरंिभक पिरणाम को प्र तुत िकया गया है।

ABSTRACT. A network of 550 Automatic Weather Stations (AWS) has been established across India during

April 2009 to February 2012 to bridge observational gaps in data sparse regions and augment manned surface observational network. All AWS are equipped with sensors for parameters air temperature, relative humidity, atmospheric pressure, rainfall, wind and global solar radiation. Additional sensors for soil temperature, soil moisture, leaf temperature and leaf wetness have been provided at 127 AWS. The data loggers are programmed for hourly measurement schedule. The satellite telemetry technique called Time Division Multiple Access (TDMA) has been employed for data transmission through Data Relay Transponder (DRT) aboard geo-stationary satellite INSAT – 3A located at 93.5º E to ensure effective channel utilization. The data are received and archived in the central data receiving Earth Station established at Pune. The mean sea level pressure, dew point temperature, pressure tendency, daily maximum and minimum temperature are derived at Earth Station. The data are made available in near real time to end users in WMO code form via WMO Information System (WIS) and at www.imdaws.com.This paper presents technical details of AWS, network plan, features of telemetry system, capabilities of data receiving Earth Station and preliminary results of the performance of network.

Key words – AWS, TDMA.

1. Introduction A steering committee (Sikka et al., 2006) constituted by the Ministry of Earth Sciences, Govt. of India recommended refurbishment and establishment of various atmospheric observational networks under modernization program for India Meteorological Department (IMD). One of the recommendations of the committee was to establish

a network of 1200 AWS and 3600 Automatic Rain Gauge (ARG) stations across India in a phased manner. Initially the AWS network would serve as augmentation of the conventional surface observatory network. Based on the results of network performance, a decision would be taken on installation of additional AWS co-located with conventional manned observatories where measurements of meteorological parameters such as air temperature,

(393)

mailto:[email protected]://www.imdaws.com/

394 MAUSAM, 65, 3 (July 2014)

Fig. 1. Network of 423 AWS

relative humidity, wind, rainfall, atmospheric pressure, solar radiation will be taken by AWS and visual observations such as visibility, present weather, cloud cover, height of base of cloud etc by human observers. Sequel to recommendations of the committee, a network of 550 AWS has been established across the country during the period April 2009 to February 2012 under IMD modernization program Phase - I. The committee made several detailed recommendations for establishment of the network of AWS. All recommendations of the committee could not be accomplished in Phase – I of modernization program. The present network plan is primarily based on three factors, viz., (i) meteorologically unrepresented districts (ii) monitoring of meso-scale systems and (iii) Agro-meteorological requirements.

Sikka et al. (2006) have mentioned that by the year 2006, there were 253 meteorologically unrepresented districts in Indo-Gangetic Plains, Himachal Pradesh, Central India and south peninsular India. These districts have been considered on priority for installation of AWS. A uniform network of AWS is now available in India with an AWS in almost all districts of the country. The present manned surface observational network is adequate for monitoring synoptic scale weather phenomena but not meso-scale severe weather generating systems. In order to address this issue for National Capital Region (NCR), a meso-network of 27 AWS has been established in and around NCR. The data of meso-network are being ingested in Numerical Weather Prediction models for generation of forecast with improved skill.

The sensors for parameters air temperature, relative humidity, atmospheric pressure, wind speed, wind

RANALKAR et al. : NETWORK OF AWS : TDMA TYPE 395

Fig. 2. Network of 127 Agro-AWS in different Agro-climatic zones of India

direction, rainfall and global solar radiation have been interfaced with each AWS. For better understanding of the processes affecting crop growth, crop yield, incidences of pests and diseases, additional sensors for soil temperature, soil moisture, leaf temperature and leaf wetness have been interfaced with 127 AWS in different agro-climatic zones of India. These stations are called Agro-AWS. The network of 423 AWS (without sensors for agro-meteorological parameters) is depicted in Fig. 1 and network of 127 Agro-AWS is shown in Fig. 2. Time Division Multiple Access (TDMA) technique is used for transmission of data through Data Relay Transponder (DRT) aboard geo-stationary satellite INSAT 3A. In this paper, we describe technical features of the network, telemetry system, Quality Control system, capabilities of data receiving Earth Station and performance of network.

2. System description and measurement schema The remote field station consists of a data logger, UHF transmitter, meteorological sensors, crossed YAGI antenna, GPS antenna, 12 V, 65 AH Sealed Maintenance Free (SMF) battery, 30 Watt solar panel, Enclosure complying to National Electrical Manufacture’s Association – 4 (NEMA – 4 ) standard and 10 m tower. The AWS system design is compact, modular, rugged and cost-effective. IMD does not own most of the sites at which stations have been installed but all efforts have been made to provide sites with exposure conditions stipulated in WMO (2008). The stations have been installed in the premises of research institutes, agricultural universities, district collector offices, colleges and schools etc. Though, not all sites comply with the stringent requirements of WMO, in general, sites have dimension of 12 m × 15 m and exposure conditions are good. The

396 MAUSAM, 65, 3 (July 2014)

TABLE 1

Characteristics of sensors used in AWS network

Parameter Sensor description Height at which

sensor is installed Make Model Range Accuracy

Air temperature Pt-100 1.2 to 1.5 m R.M. Young, USA 41382 VC -40 °C to + 60 °C ± 0.2 ºC

Relative humidity Capacitive type 1.2 to 1.5 m R.M. Young, USA 41382 VC 0 to 100 % ± 2 %

Atmospheric pressure

Solid state 1 m R.M. Young, USA 61204V-72 600 to 1100 hPa ± 0.2 hPa rms

Rainfall Tipping Bucket Rain Gauge

0.6 m Komoline, India KDS-071 Not limited by the sensor

Rain rate < 120 tips/hr

Wind speed Ultrasonic sensor 10 m Gill Instruments, UK Wind Sonic-I 0 to 60 m/s ± 2%

Wind direction Ultrasonic sensor 10 m Gill Instruments, UK Wind Sonic-I 0° to 359° ± 3°

Sunshine duration Silicon Photodiode type pyranometer

1 m LI-COR, USA 200 SZ 280 nm to 2800 nm ± 5% typical under natural daylight

conditions

Soil temperature RTD - 5 cm & -20 cm Komoline, India KDS-031 -40 °C to + 60 °C 0.1 °C

Soil moisture Time Domain Reflectometry type

-20 cm Delta-T, USA Theta Probe ML2X

Full reading range is 0 to 1 m3m-3

± 0.01 m3.m-3 (1%): 0 to 40°C

Leaf wetness Artificial leaf electrical resistance

~ 0.2 m Davis Vantage Pro2, USA

6420 Moisture level scale 0 to 15 set by the user

depending upon threshold for wetness

± 0.5

Leaf temperature Capacitive type ~0.2 m Komoline, India KDS-161 -20 °C to +60 °C ± 0.2 °C

layout of stations is consistent with that explained by Ranalkar et al. (2012). The data loggers at all AWS are pre-programmed to take measurement of meteorological parameters using interfaced sensors at an interval of an hour. The values are logged in the internal memory of the data logger, transmitted via UHF transmitter at assigned time to INSAT 3A and are received at the central data receiving Earth Station established at Pune. All sensors in the network meet the characteristics specified in WMO (2008). The sensors are either certified by IMD or traceable to National Institute of Standards and technology (NIST) standard. The radiation sensors are traceable to World Radiometric Reference (WRR). The characteristics of sensors used in the network are given in Table 1. The hourly wind speed and wind direction are obtained after taking vector average of samples taken

every second starting from three minutes prior to full hour UTC (180 samples starting from 57th minute to full hour UTC). Three-minute wind averaging is in conformity with the IMD standard being followed at all conventional synoptic observatories. Soil temperature measurement is taken at 5 cm and 20 cm depth using a single sensor having two probes. However, the probe can be installed for measurement of soil temperature at any desired depth. Leaf temperature and Leaf wetness data are logged in data logger memory but not transmitted. The air temperature samples are taken every minute for an hour (samples are kept in buffer for an hour but not logged in the memory of the data logger) and these 60 samples are used to derive hourly maximum and hourly minimum temperature. The hourly maximum and hourly minimum temperature are logged in the data logger memory and are transmitted. Tipping Bucket Rain Gauge is used to measure daily cumulative rainfall (counter is reset every day at 0300 UTC) and hourly cumulative rainfall (counter is


TABLE 2

Break-up of 230 bit data frame

Number of bits Data Sensor/

Identification code

Scale factors applied at AWS (field site) Remark

31 Bose, Chaudhuri, Hocquenghem (BCH)

code (Station IDis generated with zero appended to 31 bits)

Not applicable

Not applicable Station Identifier

5 Time (in UTC) Not applicable

Not applicable Time of measurement

11 (10 data bits + 1 parity bit)

Battery Voltage – BAT :C1 X = BAT*10 Instantaneous value of BAT in Volts measured at top of the hour


Hourly Rainfall – HR :C2 X = HR Hourly rainfall in mm (counter is reset to zero at every full hour

UTC) with value rounded off to next higher integer


Soil Moisture – SM :C3 X = SM Instantaneous value of SM in m3m-3measured at top of the hour


Health bits :H Not applicable 1 bit is used to check status of GPS lock, 1 bit is used to check sign of

temperature data

15 (10 data bits + 1 parity bit + 4 sensor identification

bits)

Sensor-I (Air Temperature – AT)

0000 (:0) X = (AT+40)*10 Instantaneous value of AT in ºC measured at top of the hour


bits)

Sensor-II (Hourly maximum Air

Temperature - ATmax)

0001 (:1) X = (ATmax+40)*10 ATmax (in ºC) during the hour (using samples taken every minute)


bits)

Sensor-III (Hourly minimum Air

Temperature - ATmin)

0010 (:2) X = (ATmin+40)*10 ATmin (in ºC) during the hour (using samples taken every minute)


bits)

Sensor-IV (Wind Speed –WSPD)

0100 (:4) X = WSPD*10.23 WSPD in m/s (using 3-minute vector averaging prior to full

hour UTC)


bits)

Sensor-V (Wind Direction – WDIR)

0101 (:5) X = WDIR*2.046 WDIR in Deg. (using 3-minute vector averaging prior to full

hour UTC)


bits)

Sensor-VI (Station Level Pressure – SLP)

0110 (:6) X = (SLP-Datum)*10.23 SLP (in hPa) measured at top of the hour (UTC)


bits)

Sensor-VII (Relative Humidity - RH)

0111 (:7) X = RH*10.23 RH (in %) measured at top of the hour (UTC)


bits)

Sensor-VIII (Cumulative Rainfall – CR)

1100 (:12) X = CR Cumulative rainfall in mm (counter is reset to zero everyday at 0300

UTC) with value rounded off to the next higher integer


bits)

Sensor-IX (Soil Temperature – ST)

1101 (:13) X = (ST+40)*10 ST (in ºC) measured at top of the hour (UTC)


bits)

Sensor-X (Global Solar Radiation)

1110 (:14) X Duration of bright sunshine in minutes is derived (counter is reset

to zero everyday at 2000 UTC)

398 MAUSAM, 65, 3 (July 2014)

Figs. 3(a-c). (a) Additive data scrambler with initial state 696916, (b) Half rate convolution encoder with constraint length 7 and (c) Format

of TDMA data burst

reset at every full hour UTC). The daily cumulative duration of bright sunshine in minutes is computed from global solar radiation sensor (counter is reset every day at 2000 UTC). The system configuration is hard-coded in the data logger and can be modified by updating the data logger firmware. 3. Time division multiple access (TDMA) telemetry

technique The network uses satellite telemetry system for transmission of data from field stations and reception at central data receiving Earth Station. The transmission technique used in this network is called Time Division Multiple Access (TDMA). A dedicated DRT on board INSAT 3A has been made available by Indian Space Research Organization (ISRO) to relay AWS data. Satellite telemetry system employed for the network is

one way communication system. Hence, it is not possible to configure the stations in the network remotely from the central data receiving Earth Station. However, it is a robust system not affected by vagaries of tropical weather. 3.1. Need for TDMA technique The ALOHA or Pseudo Random Burst Sequence (PRBS) technique has been in use in IMD for AWS data transmission since last three decades. This technique is suitable when number of AWS simultaneously sharing a common frequency channel in a randomized time of transmission mode is not too large (Abramson, 1977). As the number of AWS increases, the loss of data bursts due to collision also increases. It has been shown by Muthuramlingam et al. (2006) and Ranalkar et al. (2012) that maximum number of stations that can be transmitted in an hour using PRBS technique per satellite channel should be restricted to 400 to prevent loss of data due to burst collision.


Fig. 4. Block diagram of TDMA type AWS data receiving Earth Station

In view of massive expansion of surface observational network envisaged in near future by various organizations, need for more efficient transmission technique was felt by all DRT users. In order to address this issue Indian Space Research Organization (ISRO) recommended using Time Division Multiple Access Technique (TDMA) for transmission of data through DRTs aboard INSAT series of satellites. 3.2. Features of TDMA technique TDMA is an open loop system with timings derived from GPS receiver which is a part of AWS. Each AWS is assigned a unique one-second time stamp. The one second time frame is worked out taking into account 20 ms differential propagation delay over coverage area, RTC drift of about 1 ms per day, GPS receiver accuracy of less than 1 µsec and guard time required in receiving Earth Station. The GPS receiver in data logger updates RTC once in every 24 hours to conserve the battery. If RTC update through GPS synchronization is not achieved after 24 hours then GPS receiver in the system tries to acquire signal every hour for next three hours. If the GPS receiver fails to acquire signal even for next three hours the transmission to the satellite from particular station is suspended until GPS receiver acquires the signal. In TDMA technique, theoretically, 3600 stations can be accommodated for transmission per satellite channel. Although, not essential in TDMA technique, each AWS is configured to send repeat transmission after 30 minutes

(i.e., data burst is transmitted two times in an hour) to ensure that data is successfully received at the earth station. This reduces the channel capacity to 1800 stations, which is significantly greater than channel capacity of PRBS technique. Thus, TDMA technique is more efficient than PRBS in utilizing satellite channel capacity. It is now mandatory for all INSAT DRT users to operate future networks using TDMA transmission technique. 3.3. Description of data stream in TDMA technique The values of meteorological parameters measured by each sensor interfaced to AWS, time and satellite ID are formatted in a data frame of 230 bits. The format of data frame is same as that used in PRBS technique. A detailed break-up of data frame is given in Table 2. The data frame prefixed with Frame Synchronization (FS) code (11011000111000102, i.e., D8E216) and appended with End of Transmission (EoT) code (11111010110111102, i.e., FADE16) is subjected to Cyclic Redundancy Check (CRC). The CRC is a technique for detecting errors in data but not making corrections when data errors are detected. The CRC is calculated for 262 bits using CRC-CCITT-16 polynomial X16 + X12 + X5 + 1 and checksum bits are appended to the message after EoT. The receiver then determines whether an error occurred in transmission. The data stream comprising of FS, EoT and CRC code is then scrambled using additive scrambler defined

400 MAUSAM, 65, 3 (July 2014)

by the polynomial 1 + X-1 + X-15 of its linear feedback shift registers with initial state 695916. Data scrambler with initial state is shown in Fig. 3(a). One byte consisting of all ‘0’s is then added to the scrambled bits, after which the entire bits are convolution coded. The convolution coding is a forward error correction technique, which improves channel capacity by adding redundant information to the data being transmitted through the channel. The process of adding this redundant information is known as channel coding. The code rate (Ratio of number of bits in to the convolution encoder (k) to the number of channel symbols output by convolution encoder (n) in encoding cycle) of ½ with constraint length (number of k-bit stages that are available to feed the combinatorial logic that produces the output symbols) of 7 is used for channel coding. The octal numbers 133 and 171 represents the code generator polynomials G1 and G2. This are read as 1 + X2 + X3 + X5 + X6 and 1 + X1 + X2 + X3 + X6 and corresponds to shift register connections to modulo-2 address. The convolution encoder is shown in Fig. 3(b). The convolution code is thus obtained by combining output of k – stage shift registers through employment of n Exclusive-OR logic summers. Before convolution coding, message bits are appended with one byte consisting of all ‘0’s to ensure that every message bit proceeds entirely through the shift register and hence involved in complete coding process (Taub and Schilling, 1991). Preamble comprising of Carrier Recovery (192 symbols-all 0’s), Bit Time Recovery (64 symbols-all 1’s) and Unique Word of 64 symbols (07EA CDDA 4E2F 28C2)16 are prefixed to the convolution coded bits. The resulting 892 bits are then transmitted after differentially coded Non Return to Zero-Linear (NRZ-L) modulation at an uplink frequency of 402.74 MHz and typical transmission output power of 6-7 W with data rate of 4800 bits/sec. The duration of burst transmission is therefore 186 ms (892 bits @ 4800 bits/sec). The complete TDMA data burst format is shown in Fig. 3(c). 4. Central data receiving earth station Data transmitted from each AWS are received centrally in near real time at the data receiving Earth Station facility established at IMD Pune in the year 2009. Data bursts transmitted from AWS at assigned time stamps are received by the DRT aboard INSAT 3A at an uplink frequency. The signal is then down-converted to 28 MHz, filtered and up-converted to a downlink frequency of 4504.19 MHz. The Earth Station is capable of receiving downlink transmissions in the entire 300 MHz band of 4500-4800 MHz. The data receiving Earth Station consists of 3.8 m diameter antenna, Low Noise Amplifier (LNA) in redundant mode, extended C-band frequency down

converter in redundant mode, burst demodulator in redundant mode and data processing workstation with software. The Astra make antenna model AT512V31-SP of 3.8 m reflector diameter compatible to achieve telemetry link budget has been installed at the Earth Station. The mounting of antenna is suitable for reception of data from any INSAT satellite based DRTs located anywhere in the geostationary arc from 45° E to 115° E longitude. The antenna is used only for reception of data and can be aligned manually. Lightning and surge protection is provided to all equipments connected to antenna. The reflectors are made of solid fiberglass material. The feed mount is offset type and feed type is linear. The input frequency for feed is 4.5 to 4.8 GHz. The antenna has gain of 43 dB or more and polarization is linear. ComTech make C band LNA (model No. CLNA) with redundant switch (model RED-CLNA 1:1) is used in the Earth Station. In the event of primary LNA failure, fast automatic switch-over to the backup LNA is accomplished. The amplifiers incorporate both High Electron Mobility Transistor (HEMT) devices for low noise temperature performance and GaAs Field Effect Transistor (FET) devices for low inter-modulation. At ambient temperature of 25 °C the noise temperature is typically 45° K. The bandwidth of LNA is 300 MHz and minimum gain is 60 dB. The amplified RF signal is split at the indoor unit of the Earth Station and is fed to redundant down converters. ComTech make model DT-4503/X down-converters are used in the Earth Station system. It has +20 dBm minimum output level at the 1 dB compression point and standard gain of 45 dB. The signal is down-converted to intermediate frequency of 70 MHz. The down converted signal is fed to BPSK burst demodulator. An array of 8 demodulators is installed at the Earth Station and with redundancy 4 independent channels can be simultaneously received. The Earth Station is thus capable of receiving data from 7200 stations (1800 stations per channel). The demodulators are interfaced to servers for round the clock reception of raw data at the Earth Station. The received raw data are decoded in real time and engineering values of meteorological parameters are flushed into relational database. The dew point temperature, mean sea level pressure (for stations with elevation less than 800 m), geopotential height of nearest standard isobaric level in geopotential meter (for stations with elevation greater than or equal to 800 m), daily maximum temperature and daily minimum temperature are derived at the receiving Earth Station. After primary archival at the receiving Earth Station the hourly AWS


Fig. 5. Features of AWS metadata management software

data are passed through quality control algorithms, coded in WMO alphanumeric and BUFR form and made available to end users via WMO Information System (WIS). The hourly data are also available at www.imd.gov.in. The final archival of data are done at National Data Center of IMD, Pune after applying rigorous non real-time quality control checks. The block diagram of receiving Earth Station is shown in Fig. 4. 5. Metadata The deviations from standard exposure conditions and performance of instruments affect data quality. The end users are often interested to know the exposure conditions, type of equipment used, accuracy of sensors etc. This is particularly very important in climate studies where detailed station histories are to be considered while analyzing the data. A comprehensive metadata of each station is maintained at the receiving Earth Station using a software application. This includes geographical information of the station (latitude, longitude, elevation etc.), station description (Address, contact details of site in-charge, type of station etc.), Instruments details (make, model, serial No. of sensors and equipments, calibration details of equipment) and Station history (date of establishment, maintenance visits, occurrence and rectification of faults, replacement of equipment, relocation of stations etc.). The

features available in the metadata management software are depicted in Fig. 5. The software also has a facility to save photograph of site as an attribute of a station. After preventive maintenance of station a photograph of site can be updated to keep track of improvement / deterioration in exposure conditions. At present, the metadata software is not available in public domain. The software features are being improved to meet WMO metadata profile description (WMO, 2012) as required under WMO Information System (WIS). 6. Quality control of data The quality of AWS data depends on instruments used, site exposure conditions, installation of instruments, measurement schedule, maintenance and calibration of instruments etc. It is desirable to know the quality of data being recorded at AWS and make it known to end users. The quality of data may be affected during transmission from field station and reception at receiving Earth Station. The checks such as CRC, convolution coding, Carrier Recovery, Bit Time Recovery as explained in Section 3 are incorporated in transmission technique to ensure quality of data burst being received. The purpose of quality control is to detect errors in data. The comprehensive guidelines on quality control of surface meteorological data are given in WMO guide on the global data processing system (WMO, 1993). These

http://www.imdaws.com/

402 MAUSAM, 65, 3 (July 2014)

Fig. 6. Daily rainfall recorded on 12 November 2009 by AWS and ARG stations as cyclonic storm

‘PHYAN’ crossed the west coast

Fig. 7. Fall of station level pressure recorded by AWS installed in Mumbai as cyclonic storm ‘PHYAN’

crossed the coast near Mumbai


Fig. 8. The wind flow pattern recorded by AWS network at 1000 UTC of 31 October 2012 ahead of

landfall of cyclonic storm ‘Nilam’ (Red barb indicate station with elevation greater than 800 m)

Fig. 9. Fall of station level pressure recorded by AWS installed in Ennore Port as cyclonic storm

‘NILAM’ crossed the coast south of Chennai

404 MAUSAM, 65, 3 (July 2014)

Fig. 10. Variation of daily rainfall (in mm) on 1 November 2012 recorded by AWS installed in peninsular India

guidelines are also applicable to AWS data. A Quality Control (QC) software system for gross error check, time consistency check, climatological consistency check, internal consistency check and spatial consistency check is being developed. The first version of the software is put in operation at Earth Station. On receipt at the Earth Station, data are subjected to gross error check/range check. This check ensures that the values of all meteorological parameters are within sensor range. The software also has a provision to check the current value of parameter for climatological consistency. This feature, however, could not be implemented for all stations as many AWS are installed at new sites and hence climatology is not available. The values are then subjected to time consistency check to ensure that temporal variation of value is within acceptable limit. The software has a provision to set maximum and minimum allowable change in the value of parameter in an hour and raise appropriate flag depending upon the result of QC procedures. The algorithms for internal and spatial consistency are being developed at AWS lab. After passing the data through QC checks, the data are disseminated to end users in WMO code format. 7. Maintenance of stations Maintenance of AWS network is a challenging task, especially, for a nationwide unattended AWS network in a

tropical country like India. It is directly linked to data quality. All stations are under comprehensive warranty for a period of 24 months from the date of commissioning. On expiry of warranty, preventive and corrective maintenance of AWS is undertaken by IMD through recently constituted Regional Instruments Maintenance Center (RIMC) at each Regional Meteorological Center, viz., Delhi, Mumbai, Kolkata, Chennai, Nagpur and Guwahati. Each RIMC consists of Zonal Instruments Maintenance Centers (ZIMCs) and each ZIMC consists of Field Maintenance Units (FMUs). Biannual training programs and refresher courses are conducted in Pune at the Surface Instruments Division of IMD for all personnel involved in maintenance of AWS. Minimal set of spare sensors, maintenance tools, travelling standards, rain gauge calibrators are available at RIMCs, ZIMCs and FMUs. The preventive maintenance tours to stations once in a quarter and corrective maintenance tours as and when required should be undertaken to ensure network availability and improve quality of data. 8. Performance of the network Agnihotri (2012) compared the data of AWS in the state of Karnataka with co-located observatory for the period of May-September, 2012. The study revealed that


bias between the two time series of daily rainfall is low and features of severe thunderstorm events occurred at Bangalore were captured very well by the network. We have analyzed the performance of AWS network during two tropical cyclonic storms one in the Arabian Sea and other in the Bay of Bengal. 8.1. Case 1 : Performance of AWS network during

‘PHYAN’ cyclone In October 2009, a sufficiently dense network of 37 AWS and 65 ARG stations was available in the state of Maharashtra. The cyclonic storm 'PHYAN' crossed North Maharashtra coast between Alibag and Mumbai within 1000 UTC and 1100 UTC of 11 November 2009. Under the influence of this synoptic scale system copious rainfall occurred along the west coast and on leeward side of Western Ghat in the state of Maharashtra (IMD, 2010). The daily rainfall recorded at 0300 UTC of 12 November 2012 by AWS and ARG stations along coastal Maharashtra is depicted in Fig. 6. The fall in station level pressure as the system approached the coast was recorded by AWS installed in Mumbai and is shown in Fig. 7. The steep fall in pressure from 0400 UTC to 0800 UTC of 11 November 2009 helped in identifying landfall point. 8.2. Case 2: Performance of AWS network during

‘NILAM’ cyclone A cyclonic storm, NILAM crossed Tamilnadu coast near Mahabalipuram (south of Chennai) in the evening of 31st October, 2012 with a sustained maximum wind speed of 70-80 knots. It formed as a depression in the Southeast and adjoining Southwest Bay of Bengal at 1130 IST of 28 October 2012 near 9.5° N and 86.0° E. Subsequently, it became deep depression and then further intensified into a cyclonic storm “Nilam” in the morning of 30 October, 2012 over Southwest Bay of Bengal off Sri Lanka coast. It then moved North Northwestwards and crossed north Tamil Nadu coast near Mahabalipuram, South of Chennai between 1600 and 1700 IST of 31 October 2012 (IMD, 2012). It is reported by Mohapatra et al. (2010) that wind and pressure data recorded at AWS is very helpful in monitoring intensity and movement of landfalling cyclonic disturbances. Wind data recorded by AWS network in southern India at 1000 UTC of 31 October 2012 (half an hour before landfall) is shown in Fig. 8. Ennore Port AWS recorded maximum wind speed of 30 knot. The large scale wind flow pattern during Nilam cyclone is captured very well by the AWS network. The hourly variation of station level pressure recorded at Ennore Port from the stage of incipient cyclonic disturbance till decay after landfall is shown in Fig. 9 and

spatial distribution of daily rainfall of 1 November 2012 is shown in Fig. 10. It can be concluded from above analysis that with the help of AWS network, features of weather systems can be monitored at high temporal and spatial resolution. If the network availability and reliability is ensured through routine maintenance of network and robust near real-time quality control checks implemented at receiving Earth Station it is possible to make uninterrupted data of known quality available to forecasting centres. Acknowledgement The authors wish to express their sincere gratitude towards Dr. L. S. Rathore, Director General of Meteorology, New Delhi and Dr. Ajit Tyagi, former Director General of Meteorology for encouragement in this research work. Thanks are also due to staff at INSAT AWS Lab and all RIMCs for their dedication and hard work done during establishment of the network.

References

Abramson, N., 1977, “The Throughput of Packet Broadcasting Channels”, IEEE Transactions on Communications, COM-25, 1, 117-128.

Agnihotri, G., 2012, “Comparison of Rainfall from Ordinary and Automatic Rain Gauges in Karnataka”, TROPMET 2012 Conference, Dehradun.

India Meteorological Department, 2010, “Cyclonic disturbances over North Indian ocean during 2009: A report”, Cyclone Warning No. 5/2010 (http://www.imd.gov.in/section/nhac/dynamic/ cyclone.htm).

India Meteorological Department, 2012, “A preliminary report on cyclonic storm 'NILAM' over Bay of Bengal”, (http://www.imd.gov.in/section/nhac/dynamic/cyclone.htm)

Mohapatra, M., Kumar, Naresh and Ranalkar, M. R., 2010, “Utility of Automatic Weather Station (AWS) data for monitoring and prediction of cyclonic disturbances during 2010”, IMD Met. Monograph No. 1/2010.

Muthuramlingam, E., Kumar, Sanjay and Vashistha, R. D., 2006, “Influence of data burst collision on transmission of AWS data through satellite”, Mausam, 57, 3, 499-506.

Ranalkar, M. R., Mishra, R. P., Anjan, A. and Krishnaiah, S., 2012, “Network of Automatic Weather Stations: Pseudo Random Burst Sequence Type”, Mausam, 63, 4, 587-606.

Sikka, D. R. and Coauthors, 2006, “Atmospheric Observing Systems for India”, Internal report of the Ministry of Earth Sciences, Govt. of India.

406 MAUSAM, 65, 3 (July 2014)

Taub, H. and Schilling, D. L., 1991, “Principles of communication systems”, Tata McGraw-Hill Publishing Co., 2nd Edn., 562-563

World Meteorological Organization, 2008, “Guide to Meteorological Instruments and Methods of Observation”, Part – I, Chapter – 1.WMO No. 8, 7th Edn.

World Meteorological Organization, 2012, “Report of the second meeting of the Inter-Programme Expert Team on Metadata and Data Interoperability”, Exeter, United Kingdom.

World Meteorological Organization, 1993, “Guide on the global data processing system”, Chapter – 6, WMO No. 305, 2nd Edn.

Documents

Network of Automatic weather stations : Time …metnet.imd.gov.in/mausamdocs/16539_F.pdfSystem description and measurement schema The remote field station consists of a data logger,