EXAMINATION OF OHIO INDOOR RADON DATA
Ashok Kumar1, Akhil Kadiyala1, Vijay Devabhaktuni2, Arjun Akkala2, Dilip Varma Manthena1
Department of Civil Engineering1, Department of Electrical Engineering and Computer Science2
The University of Toledo, 2801 W. Bancroft Street, Toledo, OH 43606

Abstract
It is estimated that 700 to 1300 people die annually in Ohio from radon-induced lung cancer.
The Ohio Radon Information System (ORIS) available online at http://www.radon.utoledo.edu has
been developed to create the much-needed awareness among Ohio’s citizens on radon
concentrations in their localities and the problems associated with radon exposure. Over the
years, the database (made available by testing laboratories, Ohio Department of Health (ODH),
Ohio Environmental Protection Agency (EPA), and universities) has been expanded to 159,340
radon observations from homes; 1,341 radon observations from schools; 1,283 radon
observations from drinking water; 28,062 radon observations from licensed mitigation
contractors; and 77,581 radon observations from licensed testers and specialists. This paper
presents various statistics of the analyses performed on the radon databases developed in the last
22 years.
For various reasons, radon concentration data are not available for each and every zip code, in
Ohio. In places where data are unavailable, radon data may be estimated by applying effective
interpolation techniques allowing for comprehensive radon mitigation planning. This paper
presents two interpolation techniques for estimating radon concentration values in missing zip
codes. Initial results concerning the relative performance of such techniques are shown, and the
impact of the interpolated data on radon awareness in Ohio is discussed.

Introduction
Radon is a colorless, naturally-occurring, radioactive, inert gas formed by the natural
breakdown of uranium in soil, water and rock. Radon gas drifts upward through the ground to the
surface of the soil and seeps into the buildings through foundation cracks. Radon gas is formed
naturally by radioactive decay of uranium present in geologic materials. The major sources of
radon gas in Ohio are ‘Ohio shale’ and soil. Elevated radon levels have been discovered in every
state. The United States Environmental Protection Agency (U.S. EPA) estimates that as many as
eight million homes throughout the country have elevated levels of radon (EPA Report, 2004). If
the indoor radon concentrations exceed the EPA recommended action level of 4 pCi/l, immediate
measures should be taken to reduce the radon level to 2 pCi/l (Kumar, 2001). Radon exposure is
responsible for about 21,000 lung cancer deaths per year in the United States (EPA Report,
2003).
Ohio Department of Health (ODH) initiated an indoor radon gas program in the late 1980’s to
reduce the number of deaths attributable to radon. In the 1990s, ODH started encouraging the
reduction of radon concentrations in houses and schools to a safe level through a number of
mitigation methods. In 2001, Ohio passed a law that required radon mitigation contractors to




77

report mitigation data on homes to the ODH (ODH Chapter, 2001). As a result of the new law,
two new databases were developed to study radon mitigation systems and to track observations
by testers.
The University of Toledo (UT), under several research grants from the ODH and the Ohio Air
Quality Development Authority (OAQDA), developed the Ohio Radon Information Systems
(ORIS), and a website (Harrell, 1993; Harrell, 1991; Kumar, 2001; Kumar, 1990; and Ojha,
2001). As of June 2010, the radon database developed and maintained by UT has 159,340 radon
observations from homes; 1,341 radon observations from schools; and 1,283 radon observations
from drinking water. In addition, as of June 2009, the mitigation database has been expanded to
28,062 radon observations and the tester database has been expanded to 77,581 radon
observations. The purpose of developing and maintaining the ORIS database is to analyze radon
data across the state of Ohio and produce results that help create public awareness and
understanding of the hazards of radon gas and, therefore, reduce any of its concentration levels in
places that surpass the EPA’s action limits.
This paper summarizes the key results obtained from analysis of ORIS data and compares the
performance of two different interpolation techniques that help predict radon concentrations in
unmeasured zip codes. The two interpolation techniques used in this study are kriging and
cokriging.

Methodology
The ORIS consists of five different databases or modules: home database, school database,
water database, mitigation database, and tester database.

Home Database
The home database provides information on the radon gas concentrations measured using
radon detectors in Ohio homes. The home database was initially handled using the ORACLE/MS
Access database (Joshi, 2002). However, due to the yearly increase in the number of radon
records in the homes database it is currently handled by SQL Server 7.0/MS Access and
Microsoft Excel 2007. Data for homes is provided by various organizations and radon testing
laboratories as electronic files. The raw data is processed before inclusion in the database. As of
June 2010, there are 159,340 radon data points in the homes database. Queries were built in MS
Access to analyze the radon statistics based on counties and zip codes. The statistics computed
include maximum (Max.), minimum (Min.), arithmetic mean (AM), geometric mean (GM),
standard deviation (SD), coefficient of variance (CV), median (Md), quartile 1 (Q1), and quartile
3 (Q3). Using Geographical Information Systems (GIS) software, zip code and county maps are
drawn to visually represent the radon concentrations in Ohio. These statistics and color-coded
GM maps, for the state of Ohio based on counties and zip codes, are available online on the
ORIS website. Queries were also built to identify zip codes and counties with radon GM
concentrations > 4 pCi/l and 8 pCi/l. In this study, the unmeasured zip codes or zip codes where
the data were not available were estimated using two different interpolation techniques that
include kriging and cokriging. Some studies have provided discussion on estimating radon
concentrations for missing zip codes (Kumar, 2007; Manthena, 2009; and Akkala, 2010).
School Database




78

The School database is very small compared to the home database. MS Access is used to
handle the school database. The radon measurements in schools in Ohio are provided by the
ODH. Queries such as schools tested by county; number of schools tested each year; schools
with radon GM concentrations > 4 pCi/l, > 8 pCi/l, > 20 pCi/l; percentage of rooms having radon
concentrations > 4 pCi/l in each school; and schools with more than 15 rooms and radon
concentrations > 4 pCi/l are used to analyze the data. The school statistics query provides the
number of schools tested in each county, AM, GM, SD, and variance of radon concentrations for
1,341 schools in the 63 counties where data were measured as of June 2010. The visual
representation of the county map with the percentage of school rooms with radon GM
concentrations > 4 pCi/l is drawn using GIS and is available online on the ORIS website.
Water Database
The water database contains data collected from both public water wells and private water
wells. The public water wells database was provided by the Ohio EPA while radon data for the
private water wells were obtained from various Ohio universities’ MS theses and research
programs. Queries were built in MS Access which show the zip code, county, radon
concentrations, and date of testing for both the public and private water wells. As of June 2010,
of the 1,283 water supply systems in the water database; 216 are public water wells, and the
remaining 1,067 are private water wells. It is difficult to obtain new data in this area.
Mitigation Database
Licensed mitigation contractors perform tests and submit the results to the ODH. This data is
manually entered in to an Excel spreadsheet. The data contains information of 1) license number
of the contractor; 2) name of the contact; 3) phone number of the contact; 4) address of the
contact; 5) county, zip code, city and state; 6) system type; 7) start and completion dates; 8) premitigation and post-mitigation levels; 9) quarter of the year; and 10) year in which measurements
were made. The missing data values are assigned as “NA” for alpha numeric type and “-1” for
numeric values.
The mitigation data (Kumar, 2003) is imported to MS Access to analyze, store, and update the
mitigation data. Currently (June 2009), the mitigation database consists of 28,062 radon
observations. Queries were designed in MS Access for analyzing mitigation data. These queries
determine the total number of tests performed by each licensed contractor, average removal
efficiency by each type of system, counties with pre-mitigation radon level > 4 pCi/l, premitigation radon level between 4 pCi/l and 20 pCi/l, and pre-mitigation level greater than 20
pCi/l.
Tester Database
The radon tests conducted by licensed testers are submitted to the ODH and these records are
then passed to the UT Civil Engineering Department. As of June 2009, the tester database has
been expanded to 77,581 radon observations. The tester database includes information on 1)
license number of the tester; 2) name of the contact; 3) street address of the contact; 4) zip code,
city, county, and state; 5) device code; 6) test type; 7) start and finish dates of the test; 8) radon
concentration level; 9) quarter of the year; and 10) and year in which the measurements were
made.





79

Database maintenance procedures adopted for the tester database are similar to the
maintenance procedures adopted for the mitigation database. Tester results are entered manually
into an Excel spreadsheet, processed, and checked for accuracy to avoid transcription errors. The
data is then imported to MS Access to run queries. Queries are run to determine the number of
records tested by each licensed tester, counties with radon level > 4 pCi/l, radon statistics that
include the number of radon measurements, Max., Min., AM, GM, SD, and variance based on
county and zip codes. The queries are built in such a way that they do not consider missing data
during the analysis.

Interpolation Techniques
Radon data is not available in each and every zip code due to inapproachability, cost
effectiveness, and time constraints. Two interpolation techniques: kriging and cokriging are
investigated in this study to estimate the radon concentrations for unmeasured zip codes. These
interpolation techniques use the available radon data in known locations to estimate the radon
data for unmeasured zip codes which will help to render an effective plan to mitigate the radon
concentrations in Ohio. An overview of the two interpolation techniques is given:
Kriging: Kriging is a geostatistical technique, generally used to interpolate the value of a
random field (e.g., the elevation, z, of the landscape as a function of the geographic location) at
an unobserved location from values at observed locations. This method not only produces
prediction surface, but also provides an error and uncertainty surfaces. Kriging is mainly divided
into two different functions: predicting and quantifying the spatial structure of the data. This
interpolation technique is very flexible and allows the user to investigate graphs of spatial
autocorrelation. This technique uses statistical models that allow a variety of map outputs
including predictions, prediction standard error, standard error of indicators, and probability.
Cokriging: Cokriging interpolation technique is similar to kriging that perfoms better estimates
using a secondary variate, sampled more intensely than the primary variate. If the primary variate
is difficult or expensive to measure, cokriging uses secondary variate to predict the data without
having to more intensely sampling the primary data. This method uses radon as the primary
variate and uranium as the secondary variate. Ordinary cokriging technique has been found to
give the most reproducible estimations (Ahmed, 1987).
More information about these two interpolation techniques can be found in the open literature.
Our current research is on the use of Artificial Neural Networks to predict radon concentrations
in unmeasured zip codes (Akkala, 2010 and Akkala, 2011).

Results and Discussion
The analysis of radon data from the five databases help in better understanding the radon
problem in Ohio. The analysis of radon data and results obtained from running the queries
provide information to the concerned authorities to take necessary steps in evaluating various
steps to mitigate radon concentrations to acceptable levels (EPA action limit of 4 pCi/l).
Home Database
The analysis of home database showed that out of 88 counties in Ohio, 29 have radon GM
levels more than 4 pCi/l, with Licking being the only county with radon GM level greater than 8




80

pCi/l. Of the 1544 zip codes homes data, 32.64% zip codes have radon GM levels greater than 4
pCi/l and 8.04% zip codes have radon GM levels greater than 8 pCi/l. Based on the 159,340
radon homes database, the GM of radon concentrations in the state of Ohio is 3.99 pCi/l.
Maximum radon concentration of 927.6 pCi/l is accounted for the zip code 43952 in “Jefferson”
county. Figure 1 provides the visual representation of radon GM concentrations in Ohio on a
county and zip code basis respectively, thereby giving a better idea on the radon distribution in
Ohio.

Figure 1: Geometric Mean of Radon Concentration in Ohio Counties and Zip Codes
School Database
Sixty three (63) counties and 1,341 schools have been tested under the ODH School Testing
Program as of June 2010. Analysis of the radon data in schools on a county basis revealed a
school in Belmont County having a maximum radon concentration greater than 85.5 pCi/l. The
school also had 11 out of 39 school rooms tested with radon concentration greater than 4 pCi/l.
Figure 2 provides the visual representation of counties in Ohio with percentage of school rooms
over 4 pCi/l. It can be observed from Figure 2 that Pike County schools have approximately 5060% of schools with radon concentrations over 4 pCi/l. Overall 28% of the schools in Ohio have
a potential for at least one room in excess of the EPA action level of 4 pCi/l.
Water Database
Table 1 provides a summary of the results associated with radon concentrations in private
wells. Of the 1,067 private water wells, 65 wells have radon concentration greater than or equal
to 1000 pCi/l. Of these 65 wells, 28 are located in Logan County and 13 in Delaware County
(Table 1). Seven private wells showed radon concentration over 3000 pCi/l. However, none of
the public water supply systems have radon levels more than 1500 pCi/l. Only two public wells
showed radon concentration greater than 1000 pCi/l.




81

Figure 2: Counties with % of School Rooms with Radon Concentration >= 4 pCi/l
Mitigation Database
Mitigation testing program results provide an insight into the effectiveness of the program in
Ohio. As of June 2009, of the 28,062 radon mitigation observations; 21,607 (76.99%) records are
complete with both post-mitigation and pre-mitigation levels. Over the years, there has been an
increase in the percentage of complete records submitted. Currently, the percentage of complete
records is 94%, vs. 20.38% in 2001. This shows that the efforts of the ODH are productive which
increased data quality.
Table 2 presents the average removal efficiency by each type of system for the year 2008. The
analysis of the mitigation systems revealed that the Sub Slab Depressurization (SSD) system
performs well in mitigating the radon concentrations to below 4 pCi/l (refer Table 2). Figure 3
shows the variation of removal efficiency with pre-mitigation and post-mitigation levels for the
best mitigation system for the year 2008. It can be observed that efficiency of the system
decreases with a decrease in the pre-mitigation level, because it is difficult to reduce the radon
concentrations below a certain level. Figure 3 also shows that the removal efficiency decreases
with an increase in the post-mitigation level.
Tester Database
The query results obtained from the tester database provide an insight into the radon
concentration levels in different counties and zip codes across Ohio. Of the 77,581 records
reported as of June 2009; 77,123 records (99.41%) are complete. Table 3 provides the statistics
for counties with GM radon levels >= 4 pCi/l. There are 29 counties that have radon test levels
above 4 pCi/l. Harrison (GM = 10.25 pCi/l) and Perry (GM = 8.65 pCi/l) are the two counties
having GM radon levels > 8 pCi/l.
Table 1: Radon Concentration in Private Water Wells in Ohio Counties




82

Counties

Total No.
of Wells

Max. Radon
Conc. (pCi/l)

Min. Radon
Conc. (pCi/l)

Butler
Champaign
Clark
Clermont
Crawford
Darke
Delaware
Erie
Fulton
Greene
Hamilton
Hancock
Hardin
Henry
Huron
Logan
Marion
Miami
Montgomery
Morrow
Ottawa
Paulding
Preble
Sandusky
Seneca
Union
Warren
Williams
Wood
Wyandot

7
80
8
1
78
1
60
181
3
4
2
7
49
1
149
212
74
4
6
93
2
1
3
3
9
5
4
8
6
6

571
1491
1386
163
1021
231
2314
3104
172
703
380
470
996
510
2010
7511
1574
413
637
3425
150
190
782
130
220
334
542
245
560
180

217
73
172
326
13
231
2
20
119
200
213
180
44
510
5
25
26
174
249
25
130
190
184
80
80
82
340
148
200
96

No. of Wells with
Avg. Radon
Radon Conc. ≥
Conc. (pCi/l)
1000 pCi/l
415.00
0
355.85
2
436.88
1
163.00
0
143.42
1
231.00
0
599.12
13
285.82
6
147.00
0
438.75
0
296.50
0
322.86
0
238.73
0
510.00
0
230.82
7
553.99
28
257.61
3
248.00
0
406.67
0
303.15
4
140.00
0
190.00
0
507.67
0
96.67
0
129.89
0
210.40
0
442.00
0
185.25
0
320.00
0
132.00
0

Table 2: Average Removal Efficiency by Each Type of System for Year 2008
Type of System

Number of Records

Average % Removal

Standard Deviation

SSD
SUMP/DTD
SSD/SMD
DTD

2706
604
574
162

82.50
85.40
86.25
86.00

13.25
9.50
9.75
14.25

SUMP/DTD/SMD

41

85.25

10.50

SUMP/SSD

38

77.00

18.00




83

SUMP VENTILATON
SSD/DTD

33
23

88.33
83.50

4.67
12.67

Figure 3: Variation of Removal Efficiency with Pre-Mitigation and Post-Mitigation Levels for
the Best Performing System for Year 2008
Table 3: Statistics for Counties with Tester Radon Concentration >= 4 pCi/l
County
Harrison
Perry
Ross
Logan
Mason
Van Wert
Erie
Carroll
Madison
Pickaway
Darke
Marion
Auglaize
Coshocton
Pike
Licking
Knox
Columbiana
Champaign
Highland
Fairfield
Preble
Guernsey

No. of
Records
12
9
16
30
2
4
131
34
37
35
17
72
13
8
164
585
138
116
24
6
274
21
4

Max.

Min.

AM

GM

SD

Variance

50.50
36.60
44.60
46.10
10.10
16.70
273.20
62.10
36.60
22.60
30.10
53.60
13.10
7.80
16.50
296.60
258.00
77.10
36.80
8.80
54.00
21.70
6.70

0.80
1.70
0.50
0.40
5.20
0.80
0.05
0.40
0.50
0.40
0.60
0.20
0.50
0.70
0.50
0.20
0.40
0.20
1.00
2.30
0.20
0.10
1.30

10.74
10.70
8.15
9.13
7.65
6.82
13.34
7.87
7.46
6.31
7.07
8.30
5.52
5.30
5.79
9.75
14.60
8.83
5.72
4.88
7.58
5.37
4.50

10.25
8.65
7.91
7.56
7.25
6.77
6.76
5.69
5.59
5.48
5.32
5.32
5.08
5.08
4.99
4.98
4.95
4.93
4.90
4.88
4.86
4.61
4.50

6.61
12.22
2.65
7.49
3.46
1.20
16.28
8.78
7.44
3.95
7.44
7.83
2.48
2.40
3.10
13.80
26.13
10.09
6.07

83.93
62.29
70.83
86.55
45.29
40.75
104.27
84.24
94.59
69.38
101.99
89.95
54.11
77.56
53.51
135.96
158.11
105.32
59.69

7.41
4.47

97.47
69.15




84

Morrow
Stark
Tuscarawas
Delaware
Franklin

29
1920
526
1797
7891

22.50
111.00
134.00
735.00
939.00

0.30
0.10
0.05
0.10
0.05

5.24
6.84
7.28
6.62
6.88

4.25
4.21
4.02
3.99
3.98

3.76
9.03
10.01
9.96
10.07

69.50
128.76
127.25
132.88
144.01

Comparative Performance of Interpolation Techniques
The radon data set available at The University of Toledo, collected from various radon testing
organizations, consists of radon data for 1075 zip codes until 2010, whereas there are 1862 zip
codes mentioned in the GIS data file of Ohio. Using kriging and cokriging interpolation
techniques, radon concentrations for unmeasured zip codes are predicted and different statistical
performance measures are computed to determine the relative performance of the two
techniques. Figure 4 presents the spatial distribution of radon concentration maps obtained on
using kriging and cokriging interpolation techniques. Both these maps exhibit similar pattern of
radon concentration distributions across the state of Ohio. Relatively high concentrations are
observed in the central and western parts of Ohio as can be seen from Figure 4. It was also
observed that these maps exhibit similar pattern to the uranium distribution in Ohio. The uranium
map can be found online on the ORIS website.

Figure 4: Spatial Distribution of Radon Concentrations in Ohio using Kriging and Cokriging
Interpolation Techniques (Respectively)
Table 4 presents the summary of different statistical performance measures used to compare
kriging and cokriging interpolation techniques. The performance measures used to compare the
two techniques are model bias (MB), normalized mean square error (NMSE), correlation
coefficient (Corr.), factor of two (Fa2), fractional bias (FB), and fractional standard deviation
(FS) (Hanna, 1991). An interpolation technique is considered to be ideal and perfect if the FB
and NMSE are equal to zero. However, no technique is perfect in making accurate predictions.
Kriging and cokriging interpolation techniques are deemed acceptable if the performance
measures meet the criteria of having (i) NMSE ≤ 0.5, (ii) -0.5 ≤ FB ≤ 0.5, and (iii) Fa2 ≥ 0.80. It
can be seen from Table 4 that both kriging and cokriging techniques meet all the three



85

requirements of NMSE, FB, and Fa2. One can observe both the techniques to over predict
(negative bias) radon concentrations. Also, the predicted results using both these techniques have
shown close correlation with the observed data as can be seen from correlation coefficient values
in Table 4. One can also observe cokriging interpolation technique is slightly better than kriging
interpolation technique, because the performance measures for cokriging technique are found to
be relatively closer to the ideal values of FB, NMSE, and Fa2.
Table 4: Performance Measures for Interpolation Techniques
Performance Measures
Observed
Kriging
Cokriging

Mean
3.42
3.53
3.49

Sigma
2.74
2.47
2.34

MB
0.00
-1.25
-1.23

NMSE
0.00
0.47
0.45

Corr.
1.00
0.876
0.883

Fa2
1.00
0.99
0.99

FB
0.00
-0.358
-0.353

FS
0.00
0.101
0.114

Percentage Change in Zip Codes Exceeding 4 pCi/l Based on Kriging and Cokriging
Interpolation Techniques
Kriging and cokriging interpolation techniques were used to predict radon concentrations in
unmeasured zip codes and these results showed that there are significant number of zip codes
that exceeded 4 pCi/l. The analysis of available zip code radon data and the interpolated radon
data for missing zip codes using kriging technique showed that 32.60% of zip codes have
concentrations above 4 pCi/l as compared to 28.68% of zip codes based on measured radon data,
while cokriging technique showed 31.90% of the zip codes to exceed EPA action limit of 4 pCi/l.
These results indicate that more mitigation work is ahead for radon planners in Ohio.

Conclusion
An integrated Ohio radon information system has been successfully compiled from the data
provided by government agencies, university researchers, and commercial testing companies.
The information available from the database is useful in assessing the extent of the radon
problems in Ohio’s homes, public water systems, and schools. It was also possible to determine
the best mitigation system to control radon gas problem in Ohio homes and identify the counties
and zip codes with radon test levels greater than EPA action limit of 4 pCi/l. The radon website
developed during the project helps in creating awareness among Ohio’s citizens on radon issue
and provides information on steps to reduce radon exposure. After predicting the radon
concentrations for unmeasured zip codes using the two interpolation techniques, it was observed
that cokriging technique showed relatively better performance than kriging technique.
Interpolated data shows that more mitigation work is ahead for many zip codes that were not
known before.

Acknowledgements
The authors are thankful for the research grants awarded by the Ohio Department of Health
(ODH) and the United States Environmental Protection Agency (US EPA) to The University of
Toledo, which made the development of such a radon gas management system possible. The
contributions of earlier investigators of the grants (Dr. Jim Harrell and Dr. Andrew G.
Heydinger, and many graduate students who worked on this project over the years) are all greatly



86

acknowledged. The authors also acknowledge the contribution of a number of staff members
from the ODH. The views expressed in this paper are those of authors.




87

References
1. Ahmed S, de Marsily G. Comparison of Geostatistical Methods for Estimating
Transmission Data on Transmitivity and Specific Capacity. Water Resources Research
1987; 23: 1717-1737.
2. Akkala A, Devabhaktuni VK, Kumar A. Interpolation Techniques and Associated
Software for Environmental Data. Environmental Progress & Sustainable Energy 2010;
29(2): 134-141.
3. Akkala A, Devabhaktuni VK, Kumar A. Spatial Interpolation Techniques for
Environmental Data: Theory to Practice, Chapter 2, Advances in Environmental
Research, Volume 10, Editor: Justin A. Daniels, Nova Science Publishers, Inc., 2011 (In
press).
4. Hanna SR, Strimaitis DG, Chang, JC. Hazard Response Modeling Uncertainty (A
Quantative Method) Volume 1: User’s Guide for Software Evaluating Hazardous Gas
Dispersion Models. Sigma, Research Corporation, Westford, MA, 1991.
5. Harrell JA, Belsito ME, Kumar A. Radon Hazards Associated with the Ohio Shale.
Environmental Geology and Water Sciences 1991; 18: 17-26.
6. Harrell JA, McKenna JP, Kumar A. Geological Controls on Indoor Radon in Ohio.
Investigation No. 144, Division of Geological Survey, Dept. of Natural Resources, State
of Ohio 1993; 36 pp.
7. Joshi A, Manne GK, Kumar A. Management of Ohio's Radon Data with MS Access/SQL
Server 7.0. Environmental Progress 2002; 21 (4): D8-D12.
8. Kumar A, Harrell JA, Heydinger A. Ohio Goes Online to Combat Indoor Radon. EM (by
A&WMA) 2001; 2: 10-12.
9. Kumar A, Heydinger A, Harrell, JA. Development of an Indoor Radon Information
System for Ohio. Final Report for Ohio Air Quality Development Authority, University
of Toledo, 1990.
10. Kumar A, Maroju S, Bhat A. Application of ArcGIS Geostatistical Analyst for
Interpolating Environmental Data from Observations. Environmental Progress 2007;
26(3): 220-225.
11. Kumar A, Tandale A, Kalapati RS, Ghose S. Management of Radon Mitigation Data in
the State of Ohio. Environmental Progress 2003; 22 (3): O5-O10.
12. Manthena DV, Kadiyala A, Kumar A. Interpolation of Radon Concentrations Using GIS
Based Kriging and Cokriging Techniques. Environmental Progress 2009; 28(4): 487-492.
13. Ohio Department of Health. Radon Licensing Program Rules and Regulations. Ohio
Administrative Code, Chapter 3701-69, Ohio Department of Health 2001.
14. Ojha S, Thomas SJ, Kumar A. Experience in Integrating Geographical Information
Systems (GIS) to an Indoor Radon Database. Environmental Progress 2001; 20(3): O7O10.
15. U.S. EPA. EPA Assessment of Risks from Radon in Homes. U.S. EPA 402-R-03-003,
U.S. EPA, Air and Radiation (6608J) Washington, D.C., 20460, 2003.
16. U.S. EPA. A Citizen's Guide to Radon: The Guide to Protecting Yourself and Your
Family from Radon. U.S. EPA 402-K-02-006, U.S. EPA, Indoor Environments Division
(6609J) Washington, D.C., 20460, 2004.





88