1
00:00:00,640 --> 00:00:02,640
Speaker 1: Evening. Coming up to the next hour, we've got that

2
00:00:02,720 --> 00:00:05,400
Speaker 1: Berg official cash rate call. Tomorrow we're going to talk

3
00:00:05,400 --> 00:00:08,160
Speaker 1: to Brad Olsen about that. Shortly New Plymouth's mayor on

4
00:00:08,240 --> 00:00:11,240
Speaker 1: his challenge to labor for how labor can actually fix

5
00:00:11,280 --> 00:00:13,440
Speaker 1: this power crisis. We're in and Jamie Mackay on the

6
00:00:13,440 --> 00:00:16,640
Speaker 1: Canterbury Amp Show at seven past six. Now, KNEEWA has

7
00:00:16,680 --> 00:00:19,279
Speaker 1: got itself a fancy new toy. The state owned Wheather

8
00:00:19,360 --> 00:00:23,200
Speaker 1: Research Organization has brought itself a twenty million dollar supercomputer. Now,

9
00:00:23,239 --> 00:00:26,080
Speaker 1: once this thing is installed in two data centers in Auckland,

10
00:00:26,320 --> 00:00:28,560
Speaker 1: it will be the biggest research computer in the country.

11
00:00:28,640 --> 00:00:31,000
Speaker 1: Rob Murdoch is niea's deputy chief executive.

12
00:00:31,000 --> 00:00:32,839
Speaker 2: Hey Rob, Hi, how are you?

13
00:00:32,920 --> 00:00:34,360
Speaker 1: I'm very well, thank you. Does this mean we're going

14
00:00:34,400 --> 00:00:36,360
Speaker 1: to get more accurate weather forecasts?

15
00:00:36,840 --> 00:00:39,800
Speaker 2: Well, hopefully we will, yes in the future, and hopefully

16
00:00:39,960 --> 00:00:44,159
Speaker 2: it'll be used for a whole range of the fruit applications.

17
00:00:44,200 --> 00:00:47,839
Speaker 2: Though we'll hopefully get better at modeling our future climate

18
00:00:47,920 --> 00:00:49,920
Speaker 2: as well as our oceans and a whole range of

19
00:00:49,960 --> 00:00:52,040
Speaker 2: other environmental Thanks.

20
00:00:52,360 --> 00:00:55,600
Speaker 1: How much more accurate, like, for example, will it actually

21
00:00:55,600 --> 00:00:57,760
Speaker 1: have if we had had it in January last year,

22
00:00:57,760 --> 00:01:00,560
Speaker 1: would it have been able to forecast those rains that

23
00:01:00,600 --> 00:01:02,280
Speaker 1: we got at the start of the year.

24
00:01:03,400 --> 00:01:06,600
Speaker 2: Well, I think the key in having a bigger and

25
00:01:06,840 --> 00:01:11,920
Speaker 2: better supercomputer means that we can actually model things at

26
00:01:11,959 --> 00:01:14,760
Speaker 2: a higher resolution both in time and space. So it

27
00:01:14,800 --> 00:01:18,160
Speaker 2: means we can model things down to smaller scales, which

28
00:01:18,200 --> 00:01:20,480
Speaker 2: is important when you have to think about things like

29
00:01:20,560 --> 00:01:24,480
Speaker 2: Cyclone Gabriel for example, where it is very you know,

30
00:01:24,480 --> 00:01:27,360
Speaker 2: it's very isolated and the hills can have a massive

31
00:01:27,400 --> 00:01:29,399
Speaker 2: impact on the amount of rainfall, and the better we

32
00:01:29,480 --> 00:01:32,720
Speaker 2: can model those hills, the best of the forecasts will be.

33
00:01:33,319 --> 00:01:35,679
Speaker 2: But it also means we can do things faster. So

34
00:01:35,760 --> 00:01:38,720
Speaker 2: instead of at the moment our weather models, for example,

35
00:01:38,800 --> 00:01:44,000
Speaker 2: our high resolution where the models can predict whether in

36
00:01:44,040 --> 00:01:47,039
Speaker 2: the location around one point five square kilometers, but we

37
00:01:47,080 --> 00:01:50,000
Speaker 2: can actually get down to three hundred meters. But to

38
00:01:50,040 --> 00:01:53,720
Speaker 2: do that we need more compute power. And what this

39
00:01:54,400 --> 00:01:57,000
Speaker 2: new system will nable us to do is to be

40
00:01:57,040 --> 00:01:59,440
Speaker 2: able to instead of having models that take six hours

41
00:01:59,440 --> 00:02:01,639
Speaker 2: to run, we'll be able to run them in three hours.

42
00:02:01,680 --> 00:02:04,960
Speaker 2: So that means instead of having four forecasts today, we

43
00:02:05,040 --> 00:02:07,960
Speaker 2: can have obviously double that number.

44
00:02:09,080 --> 00:02:11,920
Speaker 1: So taking the example of cycling, Gabrielle. Would it have

45
00:02:11,919 --> 00:02:15,840
Speaker 1: avoided the situation where the Esk Valley was called unawares

46
00:02:15,880 --> 00:02:17,839
Speaker 1: they would have been told in advance if you had

47
00:02:17,840 --> 00:02:18,680
Speaker 1: this computer at.

48
00:02:18,560 --> 00:02:23,440
Speaker 2: The time, Well, I think I suspect not. I think

49
00:02:23,480 --> 00:02:26,080
Speaker 2: we needed more in place than what we currently have.

50
00:02:26,400 --> 00:02:31,000
Speaker 2: And you know, that was a very severe event and

51
00:02:31,040 --> 00:02:33,519
Speaker 2: it was very difficult to predict, and certainly there was

52
00:02:33,560 --> 00:02:35,680
Speaker 2: a lot more rainfall, although we did tend to get

53
00:02:35,720 --> 00:02:39,040
Speaker 2: the rainfall levels somewhat right. I think the other issue

54
00:02:39,080 --> 00:02:41,120
Speaker 2: in all of this is that we're not only going

55
00:02:41,200 --> 00:02:43,360
Speaker 2: to predict the rainfall. We've actually got to predict what

56
00:02:43,400 --> 00:02:46,080
Speaker 2: happens once that water touches the ground, and in that

57
00:02:46,120 --> 00:02:48,320
Speaker 2: case it causes a flood. We've got to be able

58
00:02:48,360 --> 00:02:52,359
Speaker 2: to really have good models around flood forecasting as well,

59
00:02:52,400 --> 00:02:54,880
Speaker 2: and that's what will enable us to do.

60
00:02:55,000 --> 00:02:56,840
Speaker 1: I mean, what everybody in Auckland would love to know

61
00:02:56,960 --> 00:02:59,200
Speaker 1: is basically what I was asking you at the very start,

62
00:02:59,320 --> 00:03:01,720
Speaker 1: is could we would we have known that those that

63
00:03:01,720 --> 00:03:04,760
Speaker 1: that water hitting the ground in January last year was

64
00:03:04,800 --> 00:03:06,280
Speaker 1: going to cause the trouble it was going to cause,

65
00:03:06,280 --> 00:03:08,120
Speaker 1: both by the sounds of things, not necessarily even with

66
00:03:08,200 --> 00:03:08,720
Speaker 1: this computer.

67
00:03:10,440 --> 00:03:12,480
Speaker 2: No, it depends on the type of rainfall it is

68
00:03:12,600 --> 00:03:15,960
Speaker 2: and some some some types of rainfall are easier to

69
00:03:16,000 --> 00:03:20,120
Speaker 2: predict than others, and when it gets these flash flooding

70
00:03:20,120 --> 00:03:22,840
Speaker 2: events aren't quite so easy. But there's no doubt that

71
00:03:23,200 --> 00:03:27,200
Speaker 2: the higher compute power will enable us to better develop

72
00:03:27,480 --> 00:03:30,000
Speaker 2: the models that we've got and be able to work

73
00:03:30,040 --> 00:03:32,680
Speaker 2: at a higher resolution both in time and in space.

74
00:03:33,160 --> 00:03:36,000
Speaker 2: And yes, we should be better at forecasting these things.

75
00:03:36,040 --> 00:03:37,560
Speaker 1: Does met Service have anything like this?

76
00:03:39,160 --> 00:03:39,400
Speaker 2: No?

77
00:03:39,680 --> 00:03:42,600
Speaker 1: So are you going to put them out of business? Then? No?

78
00:03:42,800 --> 00:03:45,040
Speaker 2: Well, hopefully not. I mean there's a review at the moment,

79
00:03:45,080 --> 00:03:49,960
Speaker 2: and hopefully we'll see met Service and there's obviously we're

80
00:03:50,000 --> 00:03:53,200
Speaker 2: looking across both NEEWA and net met Service about how

81
00:03:53,200 --> 00:03:55,480
Speaker 2: we can make sure we've got a forecasting system that's

82
00:03:55,480 --> 00:03:56,840
Speaker 2: fit for purpose for the country.

83
00:03:56,920 --> 00:03:58,640
Speaker 1: Where is that review at because it was supposed to

84
00:03:58,680 --> 00:03:59,520
Speaker 1: have wrapped up in May.

85
00:04:00,960 --> 00:04:04,000
Speaker 2: Yeah, I think it's as far as I'm aware, there

86
00:04:04,040 --> 00:04:07,800
Speaker 2: has been a final report completed. So it's just we're

87
00:04:07,800 --> 00:04:08,440
Speaker 2: just waiting on.

88
00:04:08,440 --> 00:04:09,920
Speaker 1: Time now to sitting with the minister.

89
00:04:10,600 --> 00:04:13,200
Speaker 2: Sitting with the ministers and see what happens.

90
00:04:13,440 --> 00:04:17,239
Speaker 1: Are you still competing with met Service for weather forecasting contract?

91
00:04:18,360 --> 00:04:20,960
Speaker 2: I don't think we've ever competed. We work very closely

92
00:04:21,000 --> 00:04:25,320
Speaker 2: with the MET Service. We have different clients and certainly

93
00:04:25,360 --> 00:04:29,599
Speaker 2: our modeling capability enables us to deliver some products for

94
00:04:29,640 --> 00:04:31,640
Speaker 2: clients that are different to the MET Service, but we're

95
00:04:31,720 --> 00:04:32,520
Speaker 2: very complimentary.

96
00:04:32,839 --> 00:04:34,800
Speaker 1: But either way, if I have to look at a

97
00:04:34,839 --> 00:04:37,000
Speaker 1: new with forecast or a MET Service forecast, you've got

98
00:04:37,000 --> 00:04:38,880
Speaker 1: the supercomputer. So you're going to be more accurate, now,

99
00:04:38,880 --> 00:04:39,279
Speaker 1: aren't you.

100
00:04:40,640 --> 00:04:44,159
Speaker 2: Well sometimes in some considerations, some not, but I think

101
00:04:44,200 --> 00:04:47,120
Speaker 2: you know, costing US.

102
00:04:47,000 --> 00:04:49,279
Speaker 1: Twenty million bucks, the answer is, yes, we're going to

103
00:04:49,279 --> 00:04:52,039
Speaker 1: be more accurate. It's money working, yes, yes.

104
00:04:51,920 --> 00:04:56,000
Speaker 2: Yes, yes, it's definitely in advancement. And the other element

105
00:04:56,080 --> 00:04:59,080
Speaker 2: that we need to think about is that this new

106
00:04:59,080 --> 00:05:02,599
Speaker 2: supercomputer is going to enable us to apply AI to

107
00:05:02,680 --> 00:05:06,159
Speaker 2: a lot more science questions. AI is changing the way

108
00:05:06,160 --> 00:05:08,160
Speaker 2: we do science, the type of science we can do

109
00:05:08,600 --> 00:05:11,040
Speaker 2: with the sort of information that's been collected in the

110
00:05:11,160 --> 00:05:14,320
Speaker 2: vast data sets that we're getting today. It's enabling us

111
00:05:14,320 --> 00:05:16,960
Speaker 2: science to questions that we wouldn't have thought were possible before,

112
00:05:17,560 --> 00:05:19,760
Speaker 2: thanks like satellite in the jury and all those sorts

113
00:05:19,800 --> 00:05:22,159
Speaker 2: of things. So it won't be just weather and climate.

114
00:05:22,240 --> 00:05:24,080
Speaker 2: It's going to be oceans and a whole range of

115
00:05:24,080 --> 00:05:27,520
Speaker 2: other things, everything from greenhouse gas production to you know,

116
00:05:28,400 --> 00:05:30,520
Speaker 2: and it won't be just new. There's opportunity for other

117
00:05:30,560 --> 00:05:34,440
Speaker 2: science organizations to have access to this compute, things like

118
00:05:34,520 --> 00:05:37,719
Speaker 2: seismic modeling for earthquakes. The list just goes on and on.

119
00:05:37,839 --> 00:05:39,680
Speaker 1: Yeah, very cool stuff. Hey, Rob, thanks very much for

120
00:05:39,720 --> 00:05:43,400
Speaker 1: your time. Rob Murdoch Nee is Deputy chief executive for

121
00:05:43,520 --> 00:05:46,480
Speaker 1: more from Hither DU plus Yellen Drive. Listen live to

122
00:05:46,600 --> 00:05:49,640
Speaker 1: news talks. It'd be from four pm weekdays, or follow

123
00:05:49,680 --> 00:05:51,440
Speaker 1: the podcast on iHeartRadio