COMMENT

On the global trends and spread of the COVID-19 outbreak:preliminary assessment of the potential relationbetween location-specific temperature and UV index

Sachin S. Gunthe1 & Basudev Swain1& Satya S. Patra2 & Aneesh Amte3

Received: 20 March 2020 /Accepted: 1 April 2020# Springer-Verlag GmbH Germany, part of Springer Nature 2020

AbstractThe novel coronavirus, since its first outbreak in December, has, up till now, affected approximately 114,542 people across 115countries. Many international agencies are devoting efforts to enhance the understanding of the evolving COVID-19 outbreak onan international level, its influences, and preparedness. At present, COVID-19 appears to affect individuals through person-to-person means, like other commonly found cold or influenza viruses. It is widely known and acknowledged that viruses causinginfluenza peak during cold temperatures and gradually subside in the warmer temperature, owing to their seasonality. Thus,COVID-19, due to its regular flu-like symptoms, is also expected to show similar seasonality and subside as the global temper-atures rise in the northern hemisphere with the onset of spring. Despite these speculations, however, the systematic analysis in theglobal perspective of the relation between COVID-19 spread and meteorological parameters is unavailable. Here, by analyzingthe region- and city-specific affected global data and corresponding meteorological parameters, we show that there is an optimumrange of temperature and UV index strongly affecting the spread and survival of the virus, whereas precipitation, relativehumidity, cloud cover, etc. have no effect on the virus. Unavailability of pharmaceutical interventions would require greaterpreparedness and alert for the effective control of COVID-19. Under these conditions, the information provided here could bevery helpful for the global community struggling to fight this global crisis. It is, however, important to note that the informationpresented here clearly lacks any physiological evidences, which may merit further investigation. Thus, any attempt for manage-ment, implementation, and evaluation strategies responding to the crisis arising due to the COVID-19 outbreakmust not considerthe evaluation presented here as the foremost factor.

Keywords Coronavirus . COVID-19 . Temperature . UVindex . Non-physiological reaction

Introduction

As of 10th March 2020, 114,814 confirmed cases of novelcoronavirus (also called COVID-19) infection have been re-ported worldwide since its emergence in mid-December (Du

Toit 2020). Almost 70.3% are being reported from Chinaalone, followed by Italy (8.0%), South Korea (6.5%), Iran(6.2%), France (1.2%), Spain (1%), Germany (1%), theUSA (0.6%), and Japan (0.6%), constituting more than~95% of the total reported confirmed cases. In a report re-leased by theWorld Health Organization (WHO). a major goaland objective is outlined to rapidly disseminate the national(for China) and international planning on the preparedness inresponse to the increasing and ongoing outbreak of COVID-19. The virus traces its origin for the recent outbreak in China,when three samples for bronchoalveolar lavage were collectedand detailed microbiological analysis confirmed that the virushad typical features resembling the coronavirus family.COVID-19 has a transmission characteristic typical of anyother cold-causing influenza virus: human-to-human. As perthe report published by the WHO, the common symptoms

* Sachin S. Gunthes.gunthe@iitm.ac.in

1 EWRE Division, Department of Civil Engineering, Indian Instituteof Technology Madras, Chennai 600 036, India

2 Transportation Engineering Division, Department of CivilEngineering, Indian Institute of Technology Madras, Chennai 600036, India

3 Aasha Endoscopy Center, Vasant Prestige, 5 Bunglow, Shahupuri,Kolhapur 416001, India

Journal of Public Health: From Theory to Practicehttps://doi.org/10.1007/s10389-020-01279-y

include fever, dry cough, fatigue, sputum production, short-ness of breath, sore throat, headache, etc., which can be rec-ognized as normal flu symptoms.

Since the beginning of the COVID-19 outbreak, extensiveefforts have been constantly devoted to improving under-standing about the virus. As a result, this virus, previouslyunknown to many, not only garnered tremendous researchattention, but also became a household name within a spanof merely 90 days. The outbreak has also brought many majorcities in China to a standstill and resulted in the isolation of alarge number of citizens in Italy. Such a curfew-like situationcan be difficult to impose and may not yield the desired resultsunder many circumstances (Fong et al. 2020), which is evi-dent from the steps being taken by countries to contain furtherspread of the virus. At the same time, another prevalent issueis considering the COVID-19 features that are similar to com-mon influenza viruses causing colds. The most common fea-ture of common influenza viruses is seasonality, which im-plies that, during colder temperatures, influenza caused byviruses increases, which then subside with the warming ofair temperature. This assumption, at this moment, is notbacked by any robust analysis and scientific investigations.Under the circumstances of unobtainability of any pharmaceu-tical relief to treat the COVID-19 virus, specific informationabout the non-physiological causes augmenting the survivaland spread of COVID-19 can be very valuable information forpolicy-makers dealing with management and effective controlmeasures. Since the onset of the COVID-19 outbreak, a lot ofattention has been paid to understanding cross-species trans-mission (Ji et al. 2020), efforts regarding the discovery, emer-gence, clinical diagnostic, and genomic characteristics ofCOVID-19 (Phan 2020), and initial public response (Patelet al. 2020). Further, there are requirements for details relatedto the necessity of maintaining regular physical activity whileexercising precautions (Chen et al. 2020) and, more interest-ingly, even developing an accurate model to forecast the novelCOVID-19 outbreak from small datasets (Fong et al. 2020).The relation of COVID-19 with meteorological parameters,which are very location-specific where outbreaks have beenreported, however, has not been investigated. In this study, weanalyzed the number of confirmed cases up till 7thMarch 2020 and investigated location-specific outbreaks withmeteorological parameters, including temperature, relativehumidity, UV index, and precipitation.

Materials and methods

To better understand the demographic distribution of COVID-19 on the global scale, we used data that were not derived forany in situ measurements or monitoring. Instead, we capturedthe country-/region-dependent spread from alerts generated bythe Program for Monitoring Emerging Diseases (ProMED),

which is a community-based online reporting system (Yadavet al. 2020). The very fact that reports in the ProMED are fullyscreened by a group of expert members for its validity makesthe reports reliable to be used in scientific publishing (Fisheret al. 2012). We used the key word “COVID-19” with theoption enabled for subject and post; for exclusions, see alsothe ProMED-mail database. Thereafter, we selectivelyscreened the number of infected cases listed for each countryfrom 2nd February 2020, where the accumulated infectedcases were already provided. As shown in Table 1, we con-sidered the cumulative dataset for 85 locations, including thehighly affected provinces in China, South Korea, Italy, Iran,France, Germany, the USA, Spain, and Japan. Further, themeteorological data for minimum, maximum, and averagetemperature (in °C); relative humidity (in %), UV index (anindex indicating the strength of sunburn-producing UV radia-tion), precipitation (in mm), and cloud cover were obtainedfrom the website https://www.worldweatheronline.com (lastaccessed 10th March 2020, 04:30 h IST), which is dedicatedto providing global weather content and has been used byprevious researchers (Ibekwe and Ukonu 2019). The pro-viders claim to cover around 3 million sites across the globe,and the data are used by many organizations. While it is dif-ficult to independently verify the quality and accuracy of thedata, random validation for a few cities with actual measure-ments in the past have given a strong agreement between thetwo datasets. To obtain the data for a specific city, we appliedthe search criteria by narrowing on the province/region andthen querying the most affected cities in the respectiveprovince/region obtained from various other data sources.Thus, we are of the strong opinion that the meteorologicaldata presented here are not only the true representation ofthe city/place with infected cases, but also give an accurateindication of the relation between the number of confirmedcases and the meteorological parameters.

Further, for the countries with more than 500 confirmedcases, we only chose cities that constituted more than 70%of the total confirmed cases for that respective country.Thus, we scrutinized a total of 107,351 confirmed cases across85 locations.

Once the data were obtained, we prepared scatter plotsbetween the number of confirmed cases and various meteoro-logical parameters. Further, by applying various statistical fits,the dependence of virus spread on individual meteorologicalparameters was investigated, as discussed below.

Assessment

Figure 1 shows the relation between the number of confirmedcases and temperature assessed for more than 10,000 casesglobally, constituting ~85 individual locations. As can be seenfrom Fig. 1a, the data points are clearly following the log

J Public Health (Berl.): From Theory to Practice

Table1

Listo

fcountriesandprovincesin

China

considered

forthisstudyandcorrespondingmaxim

um,m

inim

um,and

averagetemperature,relativehumidity

(RH),andUVindex.The

results

were

averaged

forJanuaryandFebruary

January

Region/Province/City

Num

berof

Cases

Temp

(max)

Temp

(average)

Temp

(min)

RH

Precipitatio

nRainy

Days

Cloud

Coverage

UV-

Index

Temp

(max)

Temp

(average)

Temp

(min)

RH

Hubei

66907

97

474

157.3

2163%

416

139

66Daegu

5381

74

068

137.1

1249

19

51

62Guangdong

1349

2319

1364

42.5

1446

521

1914

73Gyeonggi

141

53

–171

90.4

955

17

4–1

69Henan

1272

105

159

205

463

86

258

Seoul

120

53

055

82.1

748

17

40

56Zhejiang

1205

86

381

97.5

2370

315

128

71Barcelona

7214

1210

65262.9

732

416

1411

75Hunan

1018

98

582

154

1968

415

139

80Álava

122

1310

677

66.9

1753

315

128

75Anhui

990

76

382

148

2272

214

117

71Madrid

578

119

668

54.2

1144

418

139

63Jiangxi

935

1110

779

143.5

2166

415

139

77LaRioja

102

86

379

59.5

1551

313

106

74Sh

andong

756

64

057

27.3

441

311

83

52Florida

1323

1912

78209.2

2850

426

2011

76Jiangsu

631

86

379

113

2373

313

116

70Texas

1316

1512

78116

2046

416

1512

77Chongquing

576

1312

964

36.1

1660

315

1310

68California

105

1310

553

42.1

164

1713

737

Sichuan

538

1311

762

1015

543

1613

960

Wahington

136

96

265

77.2

1244

210

73

64Heilongijang

480

–12

–15

–20

938.9

263

2–8

–11

–17

92New

York

106

64

161

99.9

1648

17

51

62Beijin

g413

42

–245

1.6

129

37

51

45Île-de-France

104

98

579

43.1

1557

311

97

74Massachusetts

283

1–4

7371.6

1250

23

1–4

72Sh

anghai

337

119

571

116.9

2065

314

127

66Auvergne-Rhône-A

lpes

102

97

378

40.3

1045

312

105

71Hebei

318

52

–352

7.7

138

310

61

44Hauts-de-France

173

87

483

26.6

1753

310

84

77Fu

jian

296

1815

1079

14.4

1059

418

1610

80Grand

Est

250

86

381

23.3

1244

310

84

75Guangxi

252

2018

1471

3415

665

2119

1575

Bourgogne-Franche-Com

té129

85

285

7812

513

107

480

Shaanxi

245

86

157

7.6

546

215

115

39North

Rhine-W

estphalia

484

86

383

35.6

1757

39

84

78Yunnan

174

118

270

71.1

1228

213

105

65

J Public Health (Berl.): From Theory to Practice

Tab

le1

(contin

ued)

Baden-W

uerttemberg

204

75

279

27.2

1345

39

73

74Hainan

168

2222

1878

37.8

2656

523

2219

80Bavaria

256

64

076

41.6

1046

38

62

71Guizhou

146

119

582

96.2

2071

213

117

83Mazandaran

633

1210

567

7114

434

1411

668

Tianjin

136

52

–246

2.3

233

38

51

51Gilan

524

1210

777

291

2156

413

106

72Sh

anxi

133

10

–467

24.7

139

28

50

43Qom

712

97

359

35.7

1135

213

105

47Liaoning

122

–5–7

–12

803.9

132

20

–2–7

66Esfahan

601

96

249

7.4

521

313

104

37HongKong

9520

1916

7421.6

1547

520

1916

78Markazi

389

75

251

25.6

1035

210

84

45Jilin

93–9

–13

–18

909.5

352

2–6

–8–13

78Tehran

1945

75

153

2511

323

107

248

Gansu

912

–1–5

665.6

430

27

4–2

39Milan(Lom

bardia)

506

97

569

55.5

635

314

117

58Xinjiang

76–5

–8–12

7624

535

1–1

–3–7

68Piacenza

(EmiliaRom

agna)

602

97

382

5810

403

1310

569

InnerMongolia

75–6

–9–14

7910

138

21

–3–8

61Lodi(Lom

bardia)

928

97

478

38.3

635

313

106

67Ningxia

730

2–5

553

229

27

3–1

30Brescia(Lom

bardia)

739

97

472

25.8

534

312

106

63Taipei

3919

1713

7557

2260

420

1914

71Bergamo(Lom

bardia)

1245

97

368

35.9

530

312

105

60Qinghai

18–1

–5–11

603

728

25

–1–8

38Cremona(Lom

bardia)

916

97

481

26.9

637

313

76

68Aichi

670

1110

759

6820

463

119

656

Switzerland

268

84

–177

38.6

1136

29

61

77Hokkaidō

4189

119

654

95.3

1451

213

107

48UK

209

98

580

67.2

1662

310

85

70To

kyo

1180

119

554

106.7

1749

213

107

47Netherlands

188

87

485

80.2

2164

29

75

77Busan

979

74

61108.2

1348

310

84

58Belgium

169

86

383

36.8

1754

39

74

78Gyeongbuk

1081

85

271

95.8

945

29

73

67Sw

eden

161

54

079

31.7

1855

14

30

75Chungnam

987

40

7261.7

1053

19

61

70Norway

156

21

–585

47.3

651

12

1–3

74Singapore

138

2928

2679

146.5

2756

730

2826

77Malaysia

9333

3023

74192.7

3051

634

3023

72Bahrain

8519

1815

6316.6

718

520

1917

64Austria

794

30

7718.2

744

39

74

64Australia

6326

2421

77196.8

2455

725

2320

75Kuw

ait

6118

1614

510

013

519

1715

49Thailand

5035

3228

509.8

1424

735

3227

51

J Public Health (Berl.): From Theory to Practice

Tab

le1

(contin

ued)

February

Average

Region/Province/City

Precipitatio

nRainy

Days

Cloud

Coverage

UV-

Index

Temp(m

ax)

Temp(average)

Temp(m

in)

RH

Precipitatio

nRainy

Days

Cloud

Coverage

UV-

Index

Hubei

77.5

1351

312.5

106.5

70117.4

1725.815

3.5

Daegu

52.9

937

38

4.5

0.5

6595

10.5

432

Guangdong

227

1660

522

1913.5

68.5

134.75

1553

5

Gyeonggi

106.2

741

36

3.5

–170

98.3

848

2

Henan

31.9

751

49

5.5

1.5

58.5

25.95

648.5

3.5

Seoul

70.4

937

36

3.5

055.5

76.25

842.5

2

Zhejiang

103.9

1651

311.5

95.5

76100.7

19.5

60.5

3

Barcelona

2.2

622

415

1310.5

70132.55

6.5

274

Hunan

231.1

1867

312

10.5

781

192.55

18.5

67.5

3.5

Álava

47.7

1237

314

117

7657.3

14.5

453

Anhui

46.5

1245

310.5

8.5

576.5

97.25

1758.5

2.5

Madrid

0.8

319

214.5

117.5

65.5

27.5

731.5

3

Jiangxi

182

1658

313

11.5

878

162.75

18.5

623.5

LaRioja

9.6

933

310.5

84.5

76.5

34.55

1242

3

Shandong

25.6

635

48.5

61.5

54.5

26.45

538

3.5

Florida

275.6

2856

524.5

19.5

11.5

77242.4

2853

4.5

Jiangsu

36.7

946

210.5

8.5

4.5

74.5

74.85

1659.5

2.5

Texas

40.4

1651

516

1512

77.5

78.2

1848.5

4.5

Chongquing

76.5

2263

414

12.5

9.5

6656.3

1961.5

3.5

California

1.3

114

515

11.5

645

2.65

1.55

154.5

Sichuan

9.2

1349

514.5

128

619.6

1451.5

4

Wahington

118.9

1654

19.5

6.5

2.5

64.5

98.05

1449

1.5

Heilongijang

26.6

464

1–10

–13

–18.5

92.5

17.75

363.5

1.5

New

York

122.8

1759

16.5

4.5

161.5

111.35

16.5

53.5

1

Beijin

g28.6

229

25.5

3.5

–0.5

4515.1

1.5

292.5

Île-de-France

157.6

2572

210

8.5

676.5

100.35

2064.5

2.5

Massachusetts

87.4

1254

13

1–4

72.5

79.5

1252

1.5

Shanghai

52.8

1045

312.5

10.5

668.5

84.85

1555

3

Auvergne-Rhône-A

lpes

55.2

1848

310.5

8.5

474.5

47.75

1446.5

3

Hebei

13.9

327

47.5

4–1

4810.8

232.5

3.5

Hauts-de-France

116.5

2571

29

7.5

480

71.55

2162

2.5

Fujian

6310

545

1815.5

1079.5

38.7

1056.5

4.5

Grand

Est

188.7

2263

39

73.5

78106

1753.5

3

J Public Health (Berl.): From Theory to Practice

Tab

le1

(contin

ued)

Guangxi

63.3

1967

520.5

18.5

14.5

7348.65

1766.5

5

Bourgogne-Franche-Com

té127.6

2063

39

63

82.5

102.8

1657

3

Shaanxi

166

305

11.5

8.5

348

11.8

5.5

383.5

North

Rhine-W

estphalia

155.3

2372

28.5

73.5

80.5

95.45

2064.5

2.5

Yunnan

39.7

1427

412

93.5

67.5

55.4

1327.5

3

Baden-W

uerttemberg

179.5

1864

28

62.5

76.5

103.35

15.5

54.5

2.5

Hainan

40.8

2359

622.5

2218.5

7939.3

24.5

57.5

5.5

Bavaria

218.4

1962

27

51

73.5

130

14.5

542.5

Guizhou

113.6

2576

412

106

82.5

104.9

22.5

73.5

3

Mazandaran

262.7

1241

413

10.5

5.5

67.5

166.85

1342

4

Tianjin

23.6

330

26.5

3.5

–0.5

48.5

12.95

2.5

31.5

2.5

Gilan

243.9

1147

412.5

106.5

74.5

267.45

1651.5

4

Shanxi

12.2

425

34.5

2.5

–255

18.45

2.5

322.5

Qom

40.6

624

411

8.5

453

38.15

8.5

29.5

3

Liaoning

33.5

139

1–2.5

–4.5

–9.5

7318.7

135.5

1.5

Esfahan

5.2

313

511

83

436.3

417

4

HongKong

45.9

1652

620

1916

7633.75

15.5

49.5

5.5

Markazi

149.2

1028

48.5

6.5

348

87.4

1031.5

3

Jilin

8.5

248

1–7.5

–10.5

–15.5

849

2.5

501.5

Tehran

164.3

1132

48.5

61.5

50.5

94.65

1132

3.5

Gansu

6.6

421

34.5

1.5

–3.5

52.5

6.1

425.5

2.5

Milan(Lom

bardia)

8.4

827

211.5

96

63.5

31.95

731

2.5

Xinjiang

17.5

533

2–3

–5.5

–9.5

7220.75

534

1.5

Piacenza(EmiliaRom

agna)

2.6

328

211

8.5

475.5

30.3

6.5

342.5

InnerMongolia

00

222

–2.5

–6–11

705

0.5

302

Lodi(Lom

bardia)

2.9

728

211

8.5

572.5

20.6

6.5

31.5

2.5

Ningxia

00

142

3.5

2.5

–342.5

1.5

121.5

2

Brescia(Lom

bardia)

3.8

528

210.5

8.5

567.5

14.8

531

2.5

Taipei

49.8

2153

519.5

1813.5

7353.4

21.5

56.5

4.5

Bergamo(Lom

bardia)

9.2

1024

210.5

8.5

464

22.55

7.5

272.5

Qinghai

33

252

2–3

–9.5

493

526.5

2

Cremona(Lom

bardia)

1.6

329

211

75

74.5

14.25

4.5

332.5

Aichi

65.6

1336

211

9.5

6.5

57.5

66.8

16.5

412.5

Switzerland

134.9

1956

38.5

50

7786.75

1546

2.5

Hokkaidō

31.4

1440

412

9.5

6.5

5163.35

1445.5

3

UK

125.8

2167

29.5

85

7596.5

18.5

64.5

2.5

Tokyo

41.4

1640

312

9.5

650.5

74.05

16.5

44.5

2.5

J Public Health (Berl.): From Theory to Practice

Tab

le1

(contin

ued)

Netherlands

173

2576

28.5

74.5

81126.6

2370

2

Busan

89.6

835

39.5

7.5

459.5

98.9

10.5

41.5

3

Belgium

158.4

2473

28.5

6.5

3.5

80.5

97.6

20.5

63.5

2.5

Gyeongbuk

59.8

1142

28.5

62.5

6977.8

1043.5

2

Sweden

43.8

1663

14.5

3.5

077

37.75

1759

1

Chungnam

29.1

1044

48

50.5

7145.4

1048.5

2.5

Norway

83.8

756

12

1–4

79.5

65.55

6.5

53.5

1

Singapore

162.7

2250

729.5

2826

78154.6

24.5

537

Malaysia

78.4

2342

833.5

3023

73135.55

26.5

46.5

7

Bahrain

00

115

19.5

18.5

1663.5

8.3

3.5

14.5

5

Austria

52.2

1552

26.5

52

70.5

35.2

1148

2.5

Australia

568

2661

625.5

23.5

20.5

76382.4

2558

6.5

Kuw

ait

108

165

18.5

16.5

14.5

505

414.5

5

Thailand

1312

218

3532

27.5

50.5

11.4

1322.5

7.5

J Public Health (Berl.): From Theory to Practice

normal fit, with very narrow distribution. The point ofWuhan,where the maximum confirmed cases are reported, also lieswithin the peak (not shown in the figure). As evident from thefigure, it clearly appears that the temperature range of 5–15 °Ccomprises almost 90% of the total confirmed cases. With verysmall σ value of the distribution, it is very apparent that thetemperature band for community infection is, indeed, verynarrow. Previous studies have clearly demonstrated that lowertemperatures are directly related to influenza virus activity(Deyle et al. 2016; Ianevski et al. 2019; Roussel et al. 2016).Thus, it is clear that increasing temperature may have someclear indication for the reduction in the number of infectedcases. Further, to understand if the diurnal variations in tem-perature, in addition to daily maximum temperature, has anyimpact on the infection among the community, we calculatedthe difference between the daily maximum andminimum tem-peratures and plotted it against the cumulative number of con-firmed cases for respective temperature difference values (Fig.1b). It is evident from the figure that the pronounced diurnal

variations with a difference of ~7 °C between the maximumand minimum temperatures have shown a consistent decreasein community infection. However, further investigations arerequired on if such a trend is also evident for warmer andrelatively colder regions.

Further, interestingly, we also found a very strong relationbetween UV index and number of confirmed cases. As shownin Fig. 2, it is evident that the number cumulative cases werefound to be highest for a UV index of 2.5 and gradually de-creased from a UV index of 3.5. For the areas where the UVindex was higher than 5, the number of confirmed infectedcases decreased further. It has been documented that highertemperature or prolonged exposure to UVC radiation lowersthe virus infectivity (Ianevski et al. 2019). This is mainlyresulting from the fact that increased temperatures andprolonged exposure to UV radiation are unable to kill theretinal pigment epithelium (RPE) layer of the cells. Warmerwinters in Europe and the USA have been associated withreduced influenza virus activities (Ballester et al. 2016). As

Fig. 1 Scatter plot of temperature and total number of confirmed cases uptill 7thMarch 2020. Temperature data have been obtained from the onlinedataset as mentioned in the ‘Materials and methods’ section and wereaveraged over January and February 2020. a Scatter plot betweenaverage daily temperature and number of confirmed cases. The line

shows the log normal fit to the data (N = 85; R2 = 0.35). b Scatter plotof the difference between the maximum and minimum daily temperaturesand accumulated number of confirmed cases. The orange line shows thelog normal fit to the data (N = 14; R2 = 0.74). All the fits are statisticallysignificant

Fig. 2 Scatter plot betweencumulative confirmed cases andUV index (refer to the ‘Materialsand methods’ section for moredetails). The daily UV index forthe months of January andFebruary was obtained from thesource mentioned in the‘Materials and methods’ section.The daily UV index data for themonths of January and Februarywere further averaged. The lineindicates the spline fit to the data(N = 14; R2 = 0.63)

J Public Health (Berl.): From Theory to Practice

shown in Fig. 3, we did not observe any strong dependencebetween the relative humidity and precipitation and commu-nity infection.

Concluding remarks

While the world is putting every possible effort into tacklingthe global epidemic situation resulting from the novelCOVID-19 outbreak, very little information is available aboutthe novel virus. Under this situation, in this study, we present-ed the possible dependence between the infection within acommunity and the temperature and UV index. By analyzingthe data up to 7th March 2020 by considering more than 80locations comprising of a number of confirmed cases andlocation-specific meteorological data, we showed that thespread of the virus is strongly dependent on the temperatureand UV index. There is a very narrow temperature band of 3–

12 °C where 90% of the total confirmed cases are agglomer-ated. As can be seen from Fig. 4, major outbreaks across theglobe are located in the temperature areas lying in the rangeconsistent with our analysis. Most importantly, increase of theUV index has also shown a decrease in the number of cases.This is evident from the fact that the global UV index mapaveraged over 5 days for representation purposes clearlyshows a relatively lower UV index for the most affected areas(Fig. 5). This, in our opinion, is a very striking feature of ourfindings and we believe that artificial UV radiation could beone of the effective ways for sterilizing built-up environmentsfor reducing spread amongst the community. While we be-lieve that we are not presenting any physiological evidencesfor the decrease in community infections of COVID-19 due toincrease in temperature and UV index, we strongly believethat this information could be very useful, particularly whenpharmaceutical interventions are not possible at this momentto deal with the global crisis. It is important, however, to note

Fig. 3 Scatter plot between meteorological parameters and number ofconfirmed cases up till 7th March 2020. a Scatter plot between relativehumidity (%) and number of confirmed cases, with the orange and redlines indicating squared and linear fit, respectively (N = 85; R2 = 0.0005

for linear fit and R2 = 0.007 for squared fit). b Same as a but forprecipitation (N = 85; R2 = 0.0002 for linear fit and R2 = 0.01 forsquared fit

Fig. 4 Global map of the averagetemperature for January andFebruary 2020, with the countriesaffected by more than 1000 casesas of 9thMarch 2020 indicated bythe black circles. Importantly,countries, regions and locationsseverely affected are in a verynarrow band of latitude

J Public Health (Berl.): From Theory to Practice

that the information provided here is very preliminary and isnot based on any physiological studies involving either con-firmed cases and/or viruses. We advise that preparedness andother measures should take this study only for informationpurposes and propose further detailed physiological studies.

Acknowledgements The authors duly acknowledge the Program forMonitoring Emerging Diseases (ProMED), the https://www.worldweatheronline.com service, and the UV global map used in thisstudy was produced with the Giovanni online data system, developedand maintained by the NASA GES DISC. SSG acknowledges Nabha S.Gunthe and Nupura S. Gunthe for the encouragement and stimulatingdiscussions during the preparation of this manuscript. SSG is thankfulto selected faculty colleagues at Dept. of Civil Engineering at IIT Madrasfor helpful suggestions.

Compliance with ethical standards Not applicable.

Conflict of interest The authors declare that they do not have any con-flict of interest.

Role of funding source Not applicable.

Ethical approval Not applicable.

References

Ballester J, Rodó X, Robine J-M, Herrmann FR (2016) European season-al mortality and influenza incidence due to winter temperature var-iability. Nat Clim Change 6:927–930. https://doi.org/10.1038/nclimate3070

Chen PJ, Mao LJ, Nassis GP, Harmer P, AinsworthBE, Li F (2020)Coronavirus disease (COVID-19): the need to maintain regularphysical activity while taking precautions. J Sport Health Sci 9:103–104. https://doi.org/10.1016/j.jshs.2020.02.001

Deyle ER,MaherMC, Hernandez RD, Basu S, Sugihara G (2016) Globalenvironmental drivers of influenza. Proc Natl Acad Sci USA 113:13081–13086. https://doi.org/10.1073/pnas.1607747113

DuToit A (2020)Outbreak of a novel coronavirus. Nat RevMicrobiol 18:123. https://doi.org/10.1038/s41579-020-0332-0

Fisher MC, Henk DA, Briggs CJ et al (2012) Emerging fungal threats toanimal, plant and ecosystem health. Nature 484:186–194. https://doi.org/10.1038/nature10947

Fong SJ, Li G, Dey N, Crespo RG, Herrera-Viedma E (2020) Finding anaccurate early forecasting model from small dataset: a case of 2019-nCoV novel coronavirus outbreak. Int J Interact Multimed ArtifIntell 6:132–140. https://doi.org/10.9781/ijimai.2020.02.002

Ianevski A, Zusinaite E, Shtaida N et al (2019) Low temperature and lowUV indexes correlated with peaks of influenza virus activity innorthern Europe during 2010–2018. Viruses 11:207. https://doi.org/10.3390/v11030207

Ibekwe PU, Ukonu BA (2019) Impact of weather conditions on atopicdermatitis prevalence in Abuja, Nigeria. J Natl Med Assoc 111:88–93. https://doi.org/10.1016/j.jnma.2018.06.005

JiW,WangW, Zhao XF, Zai JJ, Li XG (2020) Cross-species transmissionof the newly identified coronavirus 2019-nCoV. J Med Virol 92:433–440. https://doi.org/10.1002/jmv.25682

Patel A, Jernigan DB; 2019-nCoV CDC Response Team (2020) Initialpublic health response and interim clinical guidance for the 2019novel coronavirus outbreak—United States, December 31, 2019–February 4, 2020 (reprinted from Recomm Rep, vol 68 2019). AmJ Transplant 20:889–895. https://doi.org/10.1111/ajt.15805

Phan T (2020) Novel coronavirus: From discovery to clinical diagnostics.Infect Genet Evol 79:104211. https://doi.org/10.1016/j.meegid.2020.104211

RousselM, Pontier D, Cohen J-M, Lina B, Fouchet D (2016) Quantifyingthe role of weather on seasonal influenza. BMC Public Health 16:441. https://doi.org/10.1186/s12889-016-3114-x

Yadav S, Gettu N, Swain B, Kumari K, Ojha N, Gunthe SS (2020)Bioaerosol impact on crop health over India due to emerging fungaldiseases (EFDs): an important missing link. Environ Sci Pollut ResInt 27:12802–12829. https://doi.org/10.1007/s11356-020-08059-x

Publisher’s note Springer Nature remains neutral with regard to jurisdic-tional claims in published maps and institutional affiliations.

Fig. 5 Similar to Fig. 4 butrepresenting the UV index andlocations indicated by the dots.Again, a very narrow band of UVindex is clearly evident for thelocations that are severelyaffected

J Public Health (Berl.): From Theory to Practice