RealClimate logo


Technical Note: Sorry for the recent unanticipated down-time, we had to perform some necessary updates. Please let us know if you have any problems.

Os números do nível do mar do IPCC

Filed under: — stefan @ 27 March 2007 - (English)

O aumento do nível do mar publicado no novo relatório do IPCC (o Quarto Relatório de Avaliação, AR4) já tem causado confusão considerável. Muitos artigos da mídia sugerem que há boas notícias sobre a questão do nível do mar, com previsões muito menores de aumento do nível do mar comparadas às previsões do relatório anterior do IPCC (o Terceiro Relatório de Avaliação, TAR). Alguns artigos reportam que o IPCC reduziu a projeção para o aumento do nível do mar de 88 para 59 cm, enquanto outros dizem que tal projeção teria sido reduzida de 88 para 43 cm, e existem muitas outras versões também (veja “Ampla Ironia”). Tais declarações são incorretas dado que o novo valor de até 59 cm não representa sequer toda a estória. Aqui tentarei clarear o que o IPCC de fato quer dizer e como esses números são derivados. (Mas caso prefira pular os detalhes, vá direto para a crítica ou a última linha).

O que o IPCC diz?

O Sumário para Tomadores de Decisão (SPM) lançado no ultimo mês fornece a seguinte tabela de projeções para o aumento do nível do mar:

Aumento do Nível do Mar
(em metros para 2090-2099
relativo a 1980-1999)
Caso Intervalo baseado em modelo
excetuando-se rápidas mudanças
futuras no fluxo de gelo
Cenário B1 0.18 – 0.38
Cenário A1T 0.20 – 0.45
Cenário B2 0.20 – 0.43
Cenário A1B 0.21 – 0.48
Cenário A2 0.23 – 0.51
Cenário A1FI 0.26 – 0.59

É desta tabela que sai o usualmente citado intervalo de 18 a 59 cm. O texto que acompanha a tabela diz:


Projeções baseadas em modelos da elevação do nível do mar no final do século XXI (2090-2099) são mostradas na Tabela SPM-3. Para cada cenário, o ponto médio do intervalo na Tabela SPM-3 situa-se dentro de 10% da média do modelo do TAR para 2090-2099. Os intervalos são mais estreitos que no TAR principalmente devido às melhorias na informação sobre algumas incertezas nas contribuições projetadas15. {10.6}.

nota de rodapé
15: As pojeções no TAR foram feitas para 2100, enquanto que as projeções desse relatório são para 2090-2099. O TAR deveria apresentar intervalos similares aos da Tabela SPM-3 se as incertezas tivessem sido tratadas da mesma maneira.

• Os modelos atuais não incluem incertezas do feedback climático do ciclo do carbono e tão pouco incluem efeitos completos das mudanças dos fluxos das placas de gelo, dado que ainda faltam fundamentos publicados na literatura. As projeções incluem uma contribuição devido ao aumento do fluxo de gelo da Groenlândia e Antártica em taxas observadas para 1993-2003, mas tais taxas de fluxo poderiam aumentar ou diminuir no futuro. Por exemplo, se essa contribuição crescer linearmente com a mudança da temperatura média global, os intervalos superiores da elevação do nível do mar nos cenários SRES (Relatório Especial dos Cenários de Emissão do IPCC) mostrados na Tabela SPM-3 deveriam aumentar em 0.1 m a 0.2 m. Valores maiores não podem ser excluídos, mas o conhecimento desses efeitos é muito limitado para avaliar suas probabilidades ou fornecer uma melhor estimativa ou um limite superior para o aumento do nível do mar. {10.6}

• Se a forçante radiativa fosse estabilizar em 2100 em níveis estimados no cenário A1B, a expansão térmica somente levaria a um aumento do nível do mar de 0.3 a 0.8 m em 2300 (relativo a 1980–1999). A expansão térmica continuaria por muitos séculos, devido ao tempo requerido para transportar calor para o oceano profundo. {10.7}

• A contração da camada de gelo da Groenlândia é projetada a continuar contribuindo para o aumento do nível o mar após 2100. Os modelos atuais sugerem que um aumento da perda de massa de gelo com a temperatura seria mais rápido do que um ganho de massa de gelo com a precipitação, e que o balanço de massa da superfície tornaria-se negativo sob um aquecimento global médio (relativo aos valores pré-industriais) excedendo 1.9 a 4.6°C. Se um balanço negativo de massa da superfície fosse sutentado por milênios, isso levaria a uma eliminação virtualmente completa da cobertura de gelo da Groenlândia e uma contribuição resultante do aumento do nível do mar ao redor de 7 m. As temperaturas futuras correspondentes na Groenlândia são comparáveis àquelas inferidas para o último período interglacial há 125 mil anos atrás, quando as informações paleoclimáticas sugerem uma redução da extensão de gelo polar e um aumento do nível do mar de 4 a 6 m. {6.4, 10.7}

• Processos dinâmicos relacionados o fluxo de gelo não incluídos nos modelos atuais mas sugeridos por recentes observações poderia aumentar a vulnerabilidade das placas de gelo ao aquecimento, aumentando a elevação do nível do mar no futuro. A compreensão desses processos é limitada e não há consenso sobre sua magnitude. {4.6, 10.7}

• Estudos atuais de modelos globais projetam que a camada de gelo Antártica pode permanecer muito fria para um amplo derretimento superficial e espera-se um ganho de massa devido a um aumento de queda de neve. Contudo, uma perda líquida de gelo poderia ocorrer se uma descarga dinâmica de gelo dominar o balanço de massa da camada de gelo. {10.7}

• Ambas as emissões antropogênicas passadas e futuras de dióxido de carbono deverão continuar a contribuir no aquecimento e na elevação do nível do mar por mais de um milênio, por conta da escala de tempo requerida para a remoção desse gás da atmosfera. {7.3, 10.3}

(Os itens acima documentam tudo que o SPM diz sobre o futuro da elevação do nível do mar. Os números entre chaves refem-se aos capítulos do relatório completo a ser divulgado em maio.)

O que está incluso nesses números de nível do mar?

Vamos olhar como esses números são derivados. Eles são constituídos de quatro componentes: expansão térmica, geleiras e camadas de gelo (excetuando-se as capas de gelo da Groenlândia e Antártica), balanço de massa de placas de gelo superficiais, e o desbalanço dinâmico das placas de gelo.

1. Expansão térmica (água oceânica mais quente ocupa maior espaço) é computada de modelos climáticos acoplados. Esses incluem modelos de circulação oceânica e podem assim estimar onde e quão rápido o aquecimento superficial penetra nos oceanos profundos.

2. A contribuição de geleiras e camadas de gelo (não incluindo Groenlândia e Antártica), por sua vez, é computada de uma simples formulação empírica que liga a temperatura média global à perda de massa (equivalente a uma taxa de elevação do nível do mar), baseada em dados observados entre 1963 e 2003. Tal formulação considera que as geleiras desaparecem vagarosamente e conseqüentemente param de contribuir – a quantidade total de geleiras remanecente seria suficiente para elevar o nível do mar em 15-37 cm.

3. A contribuição das duas maiores coberturas de gelo é dividida em duas partes. O que é chamado de balanço de massa superficial se refere simplesmente a queda de neve menos a ablação de gelo superficial (que é o derretimento somado à sublimação). Este é computado por um modelo de balanço de massa de placa de gelo superficial, com as quantidades de queda de neve e temperaturas derivados de um modelo de alta resolução da circulação atmosférica. Este cálculo não é o mesmo dos modelos acoplados usados nas projeções de temperatura do IPCC, de modo que os resultados desse modelo são ajustados para mimetizar diferentes modelos acoplados e diferentes cenários climáticos. (Um importante detalhe: esse balanço de massa superficial inclui algumas mudanças “morosas” no fluxo de gelo, mas essa é uma pequena contribuição.)

4. Finalmente, existe um outro modo pelo qual as placas de gelo podem contribuir para a elevação do nível do mar: ao invés de derreterem na superfície, podem começar a fluir mais rapidamente. Isso vem sendo observado com freqüência ao redor das bordas da Groenlândia e Antártica em anos recentes: saídas de geleiras e rios de gelo que drenam as placas de gelo têm aumentado suas vazões. Numerosos processos contribuem para isso, incluindo a remoção de conchas de gelo (i.e., gelos que flutuam sobre a água ancoradas em ilhas ou rochas submersas) ou a erosão da base da placa de gelo por água líquida fluindo pela superfície através de falhas no gelo. Tais processos não podem ainda ser adequadamente modelados, mas as observações sugerem que eles têm contribuído com 0 – 0.7 mm/ano para a elevação do nível do mar no período 1993-2003. As projeções na dada tabela assumem que tal contribuição simplesmente se mantém constante até o fim deste século.

Por exemplo, tome o cenário A1FI – este é o mais quente e por isso define os limites superiores do intervalo do nível do mar. A “melhor” estimativa desse cenário é 28 cm para a expansão térmica, 12 cm para as geleiras e -3 cm para o balanço de massa das placas de gelo – note que o IPCC ainda assume que a Antártica ganha mais massa através desse modo do que a Groenlândia perde. Adicionado a isso há um termo de acordo com (4) simplesmente baseado na premissa de que o acelerado fluxo de gelo observado em 1993-2003 se mantém sempre constante, adicionando outros 3 cm em 2095. No total, isso totaliza até 40 cm, com uma contribuição nula das placas de gelo. (Outro ponto importante: Isso representa um pouco menos do que a estimativa central de 43 cm para o cenário A1FI que foi divulgado na mídia, tirado dos primeiros rascunhos do SPM, pois estes 43 cm não eram a soma das melhores estimativas individuais para os diferentes fatores contribuintes, mas, ao contrário, era um ponto médio do intervalo das incertezas, o qual é um pouco maior quando algumas incertezas são tomadas com valores mais altos.)

Como esses números se comparam com o relatório anterior?


Elevação do nível do mar como verificado em Church e White 2006 mostrado em vermelho até o ano de 2001, junto com os cenários do IPCC (2001) para 1990-2100. Veja a segunda figura abaixo para um zoom no período de sobreposição.

O TAR mostrou curvas de elevação de nível do mar para uma gama de cenários de emissão (mostrada na Figura acima junto com novos dados obervacionais de Church e White 2006). Essa gama foi baseada em simulações com um modelo simples (o modelo MAGICC) ajustado para mimetizar o comportamento de uma gama de diferentes modelos climáticos complexos (por exemplo em termos de diferentes sensibilidades climáticas variando de 1.7 a 4.2 ºC), combinado com equações simples para um glacial e balanços de massa de placa de gelo (“esquema graus-dias”). Este intervalo baseado em modelo é mostrado como uma banda verde (legendada como “Several models all SRES envelope” na Figura 5 original do TAR SPM) e variou de 21 a 70 cm, enquanto que a estimativa central para cada cenário de emissão é mostrada como uma linha tracejada colorida. A maior estimativa central da elevação do nível do mar foi para o cenário A1FI (cor púrpura, 49 cm).
Ainda mais, as curvas tracejadas em cinza indicam incertezas adicionais no comportamento das placas de gelo. Tais linhas foram legendadas como “All SRES envelope including land ice uncertainty” no TAR SPM e ampliou o intervalo até 88 cm, adicionando 18 cm no limite superior. É preciso procurar minuciosamente no apêndice do Capítulo 11 do TAR para encontrar o que esses 18 cm extras representam: eles incluem uma “incerteza no balanço de massa” e uma “incerteza de dinâmica de gelo”, onde o último é meramente assumido como 10% da perda de massa total computada para a placa de gelo da Groenlândia. Note que tal incerteza na dinâmica de gelo foi somente incluída para a Groenlândia mas não para a Antártica; instabilidade da Placa de Gelo Oeste da Antártica, um cenário considerado “muito improvável” no TAR, foi explicitamente não incluído no limite superior de 88 cm.

Como mencionamos em nossa postagem sobre a divulgação do SPM, seria comparar maçãs e laranjas ao dizer que o IPCC reduziu o limite superior do nível do mar de 88 cm para 59 cm, a medida em que o primeiro incluiu “a incerteza da dinâmica do gelo” (muito embora somente para a Groenlândia, pois mudanças rápidas do fluxo de gelo na Antártica foram consideradas muito improváveis para preocupar naquele tempo), enquanto que o segundo discute essa incerteza do fluxo de gelo separadamente no texo, declarando que isso poderia adicionar 10 cm, 20 cm ou ainda mais aos 59 cm da tabela.

Assim seria melhor comparar o intervalo baseado em modelo de 21 – 70 cm do TAR com o 18 – 59 cm do AR4? Mesmo isso seria comparar maçãs com laranjas. Para um, o TAR cita a elevação até o ano 2100, o AR4 até o período 2090-2099, assim faltam os últimos cinco anos (ou 5.5 anos, mas não sejamos pedantes) da elevação do nível do mar. Para 2095, a projeção do TAR reduz de 70 cm para 65 cm (a estimativa central para o cenário A1FI reduz de 49 cm para 46 cm). Também, o intervalo do TAR é um intervalo de 95% de confiança, já o intervalo AR4 é mais estreito para um intervalo de confiança de 90%. Dados os números do TAR também como intervalos de 90% remove outros 3 cm do limite superior final.

Parece complicado? Existem outras diferenças mais técnicas… mas irei poupar-lhes disso. A reunião de Paris do IPCC já discutiu o pedido de alguns delegados de fornecer uma comparação direta dos números do AR4 e do TAR, mas desistiram de fazer isso detalhadamente por ser muito complicado. O resultado foi duas declarações:

O TAR deveria ter intervalos similares aos da Tabela SPM-3 se ele tivesse tratado as incertezas da mesma maneira.

e

Para cada cenário, o ponto médio do intervalo na Tabela SPM-3 está dentro de 10% da média do modelo TAR para 2090-2099.

(Na verdade, foi dito aos delegados pelos autores do IPCC em Paris que com os novos modelos AR4, as estimativas centrais de cada cenário seriam um pouco maiores que aquelas dos velhos modelos, se os números são reportados de forma comparável.)

A última linha mostra então que os métodos têm sido significativamente melhorados (razão por detrás de todos essas mudanças metodológicas), mas a expectativa de quanto o nível do mar irá subir no século que virá não mudou muito. A maior mudança é que a dinâmica das placas de gelo parecem mais incertas agora que no tempo do TAR, que é a razão para que esta incerteza não seja mais inclusa nos intervalos citados, mas sim discutida separadamente no texto.

Crítica – Poderiam esses números subestimar a futura elevação do nível do mar?

Existem várias discussões importantes sobre os números do nível do mar.

O primeiro é o tratamento das mudanças rápidas potenciais no fluxo de gelo (item 4 da lista acima). O AR4 aponta que as placas de gelo têm recentemente perdido massa (o período de análise é 1993-2003). A Groenlândia tem contribuído com +0.14 a +0.28 mm/ano para a elevação do nível do mar sobre esse período, enquanto que para a Antártica a incerteza varia de -0.14 a +0.55 mm/ano. É observado que a perda de massa da Antártica é predominante ou inteiramente devido às recentes mudanças do fluxo de gelo. A questão então é: Quanto esse processo irá contribuir para o futuro da elevação do nível do mar? A resposta honesta é: nós não sabemos. Como o SPM declara, pelo ano 2095 poderia ser 10 cm. Ou 20 cm. Ou mais. Ou menos.

O IPCC incluiu uma suposição no ‘intervalo baseado em modelo’ dado na tabela: tal suposição toma metade da perda de massa da Groenlândia e toda a perda de massa Antártica para 1993-2003, e assume que as perdas se manteriam sempre constantes até 2100. Essa permissa na minha visão não tem embasamento científico, pois o fluxo de gelo é quase que certamente muito variável no tempo. O relatório por si só declara que tal perda de gelo seja devida a uma aceleração recente do fluxo, e que em 2005 já era bastante alta, e no futuro os números poderiam ser várias vezes maior – ou poderiam ser menores. Incluindo um número fundamentalmente deficiente no intervalo ‘baseado em modelo’ degrada estimativas muito mais confiáveis para a expansão térmica, geleiras de montanhas e balanço de massa. Ainda pior: para os números com estimativas de erro, é adicionado um número sem uma estimativa apropriada de erro (a incerteza observada para 1993-2003 é incluída, mas quem asseguraria que esta seja válida para futuras mudanças no fluxo de gelo?). E então são apresentadas somente as margens de erro combinadas – você pode notar que nenhuma estimativa central é fornecida na tabela acima. Se eu tivesse apresentado isso como um erro de cálculo numa lição de casa no primeiro semestre de física, duvido que eu conseguiria escapar disso. A delegação alemã em Paris (da qual sou membro) então sugeriu tirar a estimativa do fluxo de gelo do intervalo tabulado. Os números se tornariam um pouco menores, mas esta abordagem não mesclaria níveis muito diferentes de incerteza, e ficaria claro o que estaria incluso na tabela e o que não estaria (as mudanças de fluxo de gelo), ao invés de tentar incluir parcialmente mudanças nos fluxos de gelo. Tais mudanças teriam sido discutidas no texto – dizendo que nas taxas de 1993-2003, tal termo poderia contribuir com 3 cm em 2095, mas esse valor poderia mudar para 10 cm ou 20 cm ou mais. Todavia, não encotramos nenhum suporte para esta proposta, a qual não teria mudado a Ciència de maneira alguma, mas melhorado a claridade da apresentação.

Como está agora, devido à forma complexa e obscura da combinação dos erros, até mesmo eu não poderia dizer por quanto o limite superior de 59 cm seria reduzido se a questionável estimativa fosse removida, e uma das razões para que os autores do IPCC não adotassem nossa proposta foi a de que os números não poderiam ser calculados rapidamente.

Um segundo problema com o intervalo acima é que os modelos usados para derivar as projeções subestimam significativamente a elevação do nível do mar em tempos pretéritos. Tentamos em vão fazer isso ser mencionado no SPM, de modo que você teria que ir ao relatório principal para encontrar essa informação. O AR4 declara que para o período 1961-2003, os modelos sobre as médias fornece uma elevação de 1.2 mm/ano, enquanto que os dados mostram 1.8 mm/ano, i.e. um crescimento 50% mais rápido. E isto sem considerar a taxa de perda de placa de gelo (0.19 mm/ano) nos números ‘modelados’ nesta comparação. Se assim fosse, a discrepância seria ainda maior – os modelos de placa de gelo prevèm que as placas de gelo ganhariam massa em função do aquecimento global. A comparação parece um pouco melhor no período de 1993-2003, para o qual os modelos fornecem uma elevação de 2.6 mm/ano enquanto os dados fornecem 3.1 mm/ano. Mas de novo as estimativas de ‘modelos’ incluem uma observada perda de massa de gelo de 0.41 mm/ano enquanto os modelos de placas de gelo fornecem um ganho de massa de 0.1 mm/ano para esse período; considerando isso, a elevação observada é de novo 50% mais rápida do que as melhores estimativas de modelos para esse período. Esta subestimativa persiste dos modelos do TAR (veja Rahmstorf et al. 2007 e Figura abaixo) – isso não é uma surpresa, desde que os novos modelos dão essencialmente os mesmos resultados dos modelos antigos, como discutido acima.


Comparação dos cenários do nível do mar do IPCC 2001 (com início em 1990) e dados observados: os dados de Church e White (2006) baseiam-se primariamente em estações de medição de maré (anual em vermelho) e dados de satélite altímetro (atualizado de Cazenave e Nerem 2004, dados espaçados de 3 meses, em azul, até meados de 2006) são mostrados com suas linhas de tendência. Note que a tendência de elevação do nível do mar segue a linha tracejada mais superior dos cenários do IPCC, exatamente aquela nomeada “incluindo a incerteza de gelo terrestre”, veja a primeira figura.

Nós então vemos que o nível do mar parece estar subindo cerca de 50% mais rápido que os modelos sugerem – consistentemente para os períodos de 1961-2003 e 1993-2003, e para os modelos TAR e AR4. Isso pode ter diversas razões, e a discrepância poderia ser considerada insignificante dados os intervalos de erros das obervações e modelos. Não há provas de que os modelos subestimam a elevação o nível do mar. Mas há no mínimo uma possibilidade plausível de que os modelos possam subestimar a elevação futura.

Uma terceira questão de importância diz respeito ao feedback do ciclo do carbono. As projeções de temperatura fornecidas na tabela SPM-3 do Sumário para Tomadores de Decisão variam de 1.1 a 6.4 ºC de aquecimento e inclui o feedback do ciclo do carbono. A variação do nível do mar, contudo, é baseada em cenários que excluem esse feedback e assim variam somente até 4.5 5.2 ºC. Isso poderia facilmente ser mal interpretado, pois na tabela SPM-3 os intervalos de temperatura que incluem o feedback do ciclo do carbono são mostrados ao lado dos intervalos do nível do mar, mas esses últimos na verdade aplicam-se a um menor intervalo de temperatura. Como uma estimativa grosseira, sugiro que para um cenário de aquecimento de 6.4 ºC, da ordem de 20 15 cm deveria ser adicionado aos 59 cm para definir o limite superior do intervalo de elevação do nível do mar.

Um ponto final seria os aspectos regionais. Gerentes de planejamento de zonas costeiras precisam ter conciência que a elevação do nível do mar não será a mesma em todos os lugares. O AR4 mostra um mapa de mudanças regionais do nível do mar, o qual mostra que por exemplo a costa européia pode esperar uma elevação de 5-15 cm a mais que a média global de elevação – isso é uma média de modelo, não incluindo a incerteza do intervalo. O padrão nesse mapa é marcadamente similar ao que seria esperado de uma desaceleração da na circulação termohalina (veja Levermann et al. 2005) de modo que provavelmente a elevação seja dominada por esse efeito. Além disso, algumas áreas terrestres estão surgindo e outras desaparecendo em resposta ao final da última era glacial ou devido à processos antropogênicos locais (como o uso de águas subterrâneas), os quais os gerentes e tomadores de decisão devem também considerar.

A última linha

A principal conclusão dessa análise é que a incerteza do nível do mar não é menor agora que na época do TAR, e citar o intervalo de 18-59 cm para a elevação do nível do mar, como muitos artigos da mídia têm feito, não representa toda a estória. 59 cm não é infortunadamente o “pior caso”. Ele não inclui toda a incerteza das placas de gelo, a qual deveria adicionar 20 cm ou mais. Ele não cobre totalmente o ‘provável’ intervalo de temperatura dado no AR4 (até 6.4 ºC) – correções nesse sentido poderiam adicionar novamente cerca de 20 15 cm. Ele não considera o fato de que a elevação passada do nível do mar seja subestimada pelos modelos por razões que são pouco claras. Considerando essas questões, uma elevação do nível do mar que exceda um metro pode, no meu ponto de vista, de modo algum ser descartada. Numa análise muito diferente, baseada somente numa simples correlação da elevação do nível do mar e temperatura, eu cheguei a uma conclusão similar. Como citado nesse paper, meu ponto aqui não é que eu tenha previsto que o nível do mar será maior que o IPCC sugere, ou que as estimativas do IPCC para a elevação do nível do mar não estejam corretas. Meu ponto é que em termos de análise de risco, o intervalo de incerteza que alguém precisa considerar é na minha visão substancialmente maior que os 18-59 cm.

Um pensamento final: esta discussão tem sido sobre a elevação do nível do mar até o ano de 2095. E tal elevação não termina nesse ano, como mostra a citação do SPM no início desse artigo. Ao longo de muitos séculos, sem esforços sérios de mitigação podemos esperar muitos metros de elevação dos oceanos. O Conselho Consultivo em Mudança Global do governo alemão (elucidando: sou membro desse conselho) em seu recente relatório especial sobre oceanos tem proposto limitar a elevação do nível do mar a um máximo de um metro, como sendo uma meta a guiar a política climática. Mas isso é uma outra estória.

Atualização: Fui recém informado por um dos autores do IPCC que os cenários de intervalo de temperatura sem o feedback do ciclo do carbono varia até to 5.2 ºC, e não 4.5 ºC como pensava. Este número não é encontrado no relatório do IPCC; tentei interpretá-lo de um gráfico, mas não exato o suficiente. Minhas desculpas! Os números no texto acima devem ser corrigidos e estão marcados. -stefan

traduzido por Ivan B. T. Lima e Fernando M. Ramos


308 Responses to “Os números do nível do mar do IPCC”

  1. 1
    pete best says:

    By how much would sea levels need to rise before coastal dwellers are adversley effected I wonder?

    Incidently even with large scale warming it is doubtful that Greenland and Antartica will disapperar completeley, come 2150 or 2200 we may see 2 to 3 meters of sea level rise?

  2. 2
    Mark A. York says:

    Since this is in centimetres could you juxtapose these figures with any level of uncertainty with the 20 feet rise should Greenland and the WAIS both melt? I watched Lomborg testify in Congress and this was the one point he made of Gore’s film where he claimed it was false beyond any doubt.

    [Response: Greenland ice is good for 7 metres and the WAIS for 6 metres of sea level rise. 20 feet is about 6 metres so either ice sheet alone, or half of each, could lead to a 20 feet rise. -stefan]

  3. 3
    John L. McCormick says:

    [The main conclusion of this analysis is that sea level uncertainty is not smaller now than it was at the time of the TAR, and that quoting the 18-59 cm range of sea level rise, as many media articles have done, is not telling the full story. 59 cm is unfortunately not the "worst case".]

    Worse case, in my view, is not the high range of sea level rise….it is merely the fact the rise will continue rising and there is no possibility for the sea rise to retreat….not in centuries or longer.

  4. 4

    Yesterday afternoon (3/26/07) I attended a public workshop on “Rapid Changes in Ice Sheet Mass Balance” organized by the US Climate Change Science Program. The people giving presentations, who know ice as well as anyone, made a clear case not only that we have little understanding of ice dynamics, but that (as one of them put it) the main sources of uncertainty in the models are all in the direction of underestimation of the sensitivity of ice sheets to a temperature rise. But then the IPCC didn’t model ice dynamics anyway, they just assumed that nothing different could happen. I think it would be hard to find anyone familiar with the subject who thinks the IPCC upper bounds are a good guide for policy makers… What were they thinking of?

    [Response: One of the reasons this stuff wasn't included in detail in the IPCC report is that it is all pretty new. Anythinig included in the report has to have stood the test of time, at least a bit. The rule was anything cited had to be in press by May 2006. Many of the important papers postdate that. All this goes to show that IPCC is for the most part, conservative. That's how science works, contrary to what the "skeptics" claim.--eric]

  5. 5
    Pierre Gosselin says:

    C’mon you guys! This is funny.
    Nobody can even predict who will win the NCAA B-Ball tournament with any reasonable certainty!
    And now some number crunchers with dubious models and no real understanding yet as to how the climate really works are telling us what the climate will be like in a 100 years? There are so many events that could happen between now and then to change everything.
    But hey, if we’re going to play this game, then here’s my prediction: global temperature will be 0.28°C colder than today and sea levels will be 93.7 mm higher. I’ll spare you my 5 other scenarios. Michael Crichton is right, an informed guess is just a guess.

  6. 6
    Andreas Mueller says:

    Thank you very much Stefan for this clarification! I think the point you mentioned about the riskassessment ist very importand for people who have to decide about the security of costal homelands for the next centuries. They should always take into consideration the worst case scenario and not only what is allowed for policymakers to read. The next point I appreciate are the graphs you presented. If the measured data are really at the upper end of the past predictions – this should alert everybody who has to deal with these issues.

  7. 7
    Alexander Ac says:

    Nice article.

    George Kukla states, that increasing temperature will lead to increased snow precipitation and snow gains will offset the sea level rise – he warns of coming ice age ;-). He apparently did not change his mind since 1970-ties…

    Considering the carbon-cycle feedback, some models (e.g. Cox et al.) estimate large positive vegetation feedback (increased soil respiration, lower photosynthesis due to increased vegetation stress, increased fire frequency…) and some of the most extreme scenarios predict the CO2 concentration to be up to 980 ppm. This is what I call catastrophe…

    Hmm – 6C rise in global temperature – i think we are in trouble even with the rise of 3C…

    Mark Serreze in his latest article in Journal Science states, that complete summer ice melt in Arctica is increasingly probable – in fact it is only matter of time, when it will happen. We know, that Arctic is melting quite rapidly. Antarctica is melting much slower, if at all. My question is:

    Do we know from from paleoclimatology, that we can have a climate state, in which there is *no* summer ice in the North pole, while the South pole “is ok”?

    I think this would create a bit of thermal imbalance in the climate and also would mean large changes in the weather pattern, which would *not* be limited only in the arctic region…

    I just wonder, what we tell our children, if they will ask what were we doing, when we were aware of this possibility…

  8. 8
    Mike says:

    Off topic comment, but I do not know where else to have this question asnswred (if anyone could kindly direct me, I would appreciate it). I’ve been chatting with a denialist, who insists that a big reason he does not “believe” in anthropogenic climate change is because “ice cores going back thousands of years cannot tell temperatures or CO2 levels when the temperature was too warm for ice formation.” ie. Perhaps a sudden, rapid spike in CO2 levels and global temperature like we are now experiencing hashappened in the past, naturally, but we are not aware of it because no ice cores exist to tell us. Now I know this is a stupid argument anyway (“a 30% increase in human CO2 and a corresponding rise in temperature are not linked, because we don’t know that it can’t happen naturally”) but I was wondering if he was at least right about the ice core data. Thanks!

  9. 9
    Alexander Ac says:

    Re#1
    Dear Andreas,

    I think, that the upper limit of sea level rise should alert *everybody*, because in the end, we *all* deal with this planet ;-)
    Or where do You think the people from coastal ares will go? :-)

  10. 10
    Nick Riley says:

    Can anyone shed any more light on this press release from last June which claims that Arctic average sea level appears to be falling?

    http://www.nerc.ac.uk/press/briefings/2006/arctic.asp

  11. 11

    For what it’s worth, I’ve had an insightful flash into how to stop global warming. According to Lilo, Pudge, a fish off the coast of Kauai, controls the weather. If we can keep him in peanut butter sandwiches, everything will be under control.

  12. 12
    PHE says:

    Re 3 (Nick Riley). What’s interesting is that this is presented as ‘something strange that needs explaining’ – both by NERC and the BBC (as linked). There is no suggestion that it could indicate global warming scares are exagerated. If this study had shown that the Arctic sea level was rising more rapidly than elsewhere, this would have been headlined as further dramatic proof of impending doom. Its true that id does not disprove AGW. But it supports the case that all pieces of ‘evidence’ need to be considered with caution.

  13. 13
    Hank Roberts says:

    Nick, you can read the abstract. It’s a first effort at digging info out of data, making corrections, and modeling what they think may be happening:

    http://www.agu.org/cgi-bin/wais?ff=G21A-02

    “… to determine Arctic sea level change …. is not an easy feat, as the ice cover obstructs the view of part of the sea surface and affects the measurements in mixed ocean/sea-ice conditions. Thus considerable effort has been put in the separation of the radar returns from leads and from sea ice. Moreover, microwave radiometer measurements of wet tropospheric delay can not be used since those measurements are also affected by sea ice. The Arctic altimeter data were retracked using an OCOG retracking algorithm, and the diffuse returns from the leads and open ocean were combined with a host of instrumental corrections and geophysical models to determine instantaneous mean sea level….”

  14. 14
    Andreas Mueller says:

    Re#2
    Yes you are right!
    I would like to know the “weight” of each decimeter sea level rise regarding the consequences for people living near or far from costal regions? I suppose the first decimeter will be within current safety margins (at least I hope so!) but what will happen if the sea level will rise to the next decimeter and so far? What are the critical margins for whom? Are there any assumptions about the ability of mankind to adapt to theses changes? Are there projections what each decimeter will cost? So if 10 cm will cost nothing, 20 cm will cost 1 billion and 30 cm will cost 100 billion then we should know something more about the weight of these measures and not only look at a scalar value.

  15. 15

    Hi folks: I found this article at MSNBC and am wondering if the opinions expressed in it have been taken into account…

    “But with ever more omens foretelling the death of the ice capsâ��possibly, in some models, by the year 2040â��researchers are launching a major effort to make such a prediction.”

    http://www.msnbc.msn.com/id/17770832/site/newsweek/

  16. 16
    Todd says:

    Thanks for clearing up some of the confusion. This point actually came up at my defense last week, as some of my committee was not aware that the new sea-level estimates did not include dynamic effects from the ice sheets, and therefore could not be considered ‘lower’ than previous estimates. It points to the importance of future dynamical modeling of ice flow.

  17. 17
    cat black says:

    Thanks for diving into this topic. The clear danger in issuing an SPM document is that nobody ends up reading the full report, nor incorporating the subtle complexities of the science into policy planning (which sadly is still likely to be the case even after the full report is issued.)

    One of the ironies of all this is that we almost daily discover the limits of our knowledge regarding these processes, and find that we have erred in our thinking always on the lower end of the range of possibilities; nature seems generally to move faster and further than we guess at first. And I’m not the first to notice this. In fairness, our science has little ground proofing of theory on climate change because this is our first time having actual field observations. That said, if our predictive science is somehow missing the trajectory of the changes then that failure ought to become a topic of discussion in itself at some point.

    Though I know it is entirely nonsense, in the back of my mind I imagine waking up one morning and catching a headline in the paper stating that Greenland has experienced a sharp increase in icequakes, and getting to work and checking CNN and finding out that half the ice sheet is currently in motion and accelerating, heading for the sea, and the UN Security Council is sitting in emergency session, and Wall Street is in a panic, and then looking out the office window at San Francisco Bay and realizing, once again, that our science was behind nature, not ahead of it, and we didn’t know how far behind we’d fallen.

    Like I said, it’s all nonsense. But I would be comforted if someone could explain to me exactly what is nonsensical about it.

  18. 18

    [[Michael Crichton is right, an informed guess is just a guess. ]]

    Michael Crichton doesn’t know his guess from a hole in the ground. If the CO2 goes up as in a given scenario, and all else is equal, and the known feedbacks are in place, then the temperature will rise by so much with error bars. What is so hard about that?

  19. 19

    [["ice cores going back thousands of years cannot tell temperatures or CO2 levels when the temperature was too warm for ice formation." ]]

    The ice cores are from places where there is snow deposition every year. We don’t measure temperatures from the layers directly, we calculate it from proxies like the level of carbon dioxide in bubbles and the O16/O18 ratio.

    [Response: Correcton: the level of carbon dioxide in bubbles is never used to infer temperature. But yes, the 18O/16O ratio is one of several methods that is used to infer temperature. And yes, snow falls even in "warm" years on the Greenland summit, and the Antarctic plateau.-eric]

  20. 20
    Lynn Vincentnathan says:

    “In terms of a risk assessment, the uncertainty range that one needs to consider is in my view substantially larger than 18-59 cm…[T]his discussion has all been about sea level rise until the year 2095. Sea level rise does not end there…”

    I think these sum up the important points.

    Here’s a site for trends by state, and it doesn’t look good for mine, Texas: http://tidesandcurrents.noaa.gov/sltrends/sltrends_states.shtml?region=tx

    Then if you add in enhanced hurricanes & floods (there was a hurricane-looking downpour over Houston on March 14th, the rain going round & round, & not moving eastward very fast)….then the situation is bad indeed.

    I wonder if our state planners are aware of global warming & its impacts (I know the water planners are not, according to a recent news article).

  21. 21
    Alexander Hill says:

    Great post. But did you see that Roger Pielke on 7 February attacked you guys for “adding to the confusion” for stating that the new 59 cm should not be compared to the old 88 cm? Clearly he was wrong and just trying to score some cheap points at your expense, without bothering to check the facts properly. His strategy seems to be to attack others so that he can then present himself as “honest broker”.

  22. 22
    Aaron Lewis says:

    Why do the folks talking about ice sheet melt, talk about “global temperature”, when what affects the ice melt/ice dynamics is the temperature adjacent to the ice? Why do we talk about a global temperature increase of 1.4 or 1.6 degrees when Greenland ice pokes up into a region where the temperature has increased much more, and the ice sheets on the Antarctic Peninsula are in an area where the temperature changes over the last few years are even greater? What is the effect of a warmer North Atlantic Drift current on Greenland? What is the effect of warmer Antarctic waters on Antarctic ice that is in direct contact with that warmer water?

    My point is that seawater at -0.5C has a very different effect on fresh water ice than seawater at +0.5C, and yet that is only 1 degree of warming. Not all degrees of warming are equal, and therefore averages do not mean much. The real question is, “What is the heat content of the water in contact with the ice sheets of concern?” This is where we can get some real near term excitement in the field of ice dynamics.

  23. 23
    David B. Benson says:

    Re #14: Andreas Mueller — The first decimeter is not free for anyone living along a flat coast. The low countries, England, Germany and especially Bangladesh come to mind, but this is hardly a complete list.

    The obvious problems are storm surges and salt water intrusion into wells, but there may well be others I haven’t learned about…

  24. 24
    Hank Roberts says:

    >MSNBC, Newsweek article claims:
    “scientists have never found this phenomenon worrisome. Until this year, when Ronald Kwok of NASA’s Jet Propulsion Laboratory rang the alarm. He’d noticed that in 2005….”

    Yeah, right. People had been trying to tell us about this for a long time before 2005. Al Gore got the Navy’s archive declassified to let this info be studied in public — in the previous millenium.

    I refute Newsweek with a couple of easily found examples, there’s plenty more — thus:
    http://www.nsf.gov/od/oia/activities/cov/opp/2001/GPRA_Rpt.doc.
    and thus:
    http://research.nps.navy.mil/cgi-bin/vita.cgi?p=display_more&id=1023568034&field=pubs&title=PUBLICATIONS

  25. 25

    It is wrong to blame the journalists for getting the story wrong when it took Stefan eight screenfuls of text and diagrams to explain what the IPCC really means! It was stupid of the IPCC to combine projections of sea level with temperature when they knew that the two were not directly linked. Did they really think that the public, journalists, and sceptics would read (or even understand) the small print in the sea level column stating “Model-based range excluding future rapid dynamical changes in ice flow”? Or was this obfustication done deliberately at the request of the Bush Administration?

    Of course, most IPCC scientists don’t believe that rapid dynamical changes in ice flow can happen! They believe in the 19th century uniformitarian paradigm of Charles Lyell, where geological change is slow and steady and all is right with the world. But that is wrong. Geology changes abruptly using eruptions, earthquakes and tsunamis, not to mention anoxic events and mass extinctions.

    Climate too is a dynamical system, in which positive feedbacks can dominate. When they do, they are short lived, because the system will continue to change until negative feedbacks take over, and the system is again stable. Once stable it will remain there, until a new shock knocks it into another abrupt change with positive feedbacks again dominating.

    We think the climate is stable because we have lived in the Holocene, which is one of those stable states. Like all other interglacials it will end abruptly, but rather than returning to an ice age, this one will switch into a hothouse world because of anthropogenic greenhouse gases.

    Re #10 where Nick Riley asks about http://www.nerc.ac.uk/press/briefings/2006/arctic.asp , the answer is that the Arctic sea ice is thinning and the fresh water from the melting ice is being replaced by denser saline sea water. This causes the small fall in sea level in the Arctic where it is happening. It means that the Arctic sea ice has continued to thin by 0.1 m per year since Rothrock et al. reported in 1999. Since it was 2 m thick in 1997, simple arithmetic shows it will all be gone by 2017 if not sooner.

    Of course this conflicts with the estimates for an ice free Arctic in 2040, 2050, and 2060 quoted by the scientists. For writing this, will I recieve another ad hominem attack, where I am accused of arrogance for daring to criticise them, and of stating the obvious? For instance – [Response: I don't usually resort to sarcasm in my original response, above, but Alastair McDonald's comment must surely rank as one of the most impressive displays of know-it-all-ness I've seen yet on RealClimate. He not only knows the cause, but he knows the solution to all our global warming concerns. -eric]

    Or will I be ignored? That is their normal technique when they have no answer, but they know that they are right!

  26. 26
    James says:

    Re #8: [...ice cores going back thousands of years cannot tell temperatures or CO2 levels when the temperature was too warm for ice formation.]

    That seems pretty easy. If you’re measuring CO2 levels in the ice, you have some carbon, so you can apply C14 dating to it, no? If you do that (or use other dating techniques, the explanation of which I’ll leave to those more knowledgeable), and you find that you don’t have any gaps in the record, then his objection is answered.

  27. 27
    Ray Ladbury says:

    Re #5. Pierre, As is clear from his books, Michael Crichton does not see the value in being informed. As such, I do not trust his assessment (or that of any other anti-science hysteric) for much of anything.
    And actually, in many cases, an informed expert on basketball can indeed pick the winner more often than not. What is more, start with the opinion of experts and let a large number of “fans” weigh in (by betting), and that is how Vegas makes much of its money. (And it’s also not too dissimilar from scientific consensus, although the “fans” being scientists themselves are more professionally trained.)
    In any case, these are scientific predictions–and scientific predictions have a pretty good track record of being true. You really ought to learn about the process.
    By quoting Crichton as an authority, all you do is diminish your own.

  28. 28

    Re #8 and #26

    I see the problem now, and it is partly true that we cannot tell the temperture beyond 150,000 years ago in Greenland because there is no ice before that date. However, we have been able to go back 700,000 years in Antarctica. Since carbon dioxide is a well mixed gas, then the CO2 in the Antarctic cores tells us about the CO2 worldwide as far back as 700,000 years ago. And we can extrapolate the temperature from the Antarctic ice core to the rest of the world.

    There is other evidence that can be used to calculate the temperatures and CO2 levels before the times recorded in the ice cores, and also during the times of the ice cores. It was ocean sediment cores that proved Milankovitch was right, and fossil beaches that supported that idea. The ice cores are used because they have been proved right by other supporting evidence, and they show that the supporting evidence can be used for times before that recorded by the ice cores.

    HTH,

    Cheers, Alastair.

  29. 29
    Steve Bloom says:

    Re #s 10 and 13: This BBC article has a bit more explanation on falling Arctic sea level.

  30. 30
    Hank Roberts says:

    Steve, the AGU abstract I linked in #13 is the study on which that BBC article is based.

    All — when a news article gives a few clues (AGU, author’s names) Google Scholar will almost always find you the actual abstract, at least.

    I quoted about half the abstract in #13 — it’s a _very_ tentative conclusion for the reasons they state there. Good, difficult work trying to tease useful information out of that sort of data set. An exercise.

  31. 31
    Dave Rado says:

    Re. 10, there’s also this article about it on realclimate; and this one is also good.

  32. 32
    yartrebo says:

    How certain are we that Greenland and Antarctica will take many hundreds or thousands of years to melt, even under heavy (5C+) global warming? My intuition is that the abnormal warming of the poles will continue, so a 5C rise in global temperature would mean perhaps a 15C rise in polar temperatures, and that should be able to melt Greenland in short order. Ice sheets have strong positive feedbacks, so I wouldn’t be surprised if they melt faster and faster as time passes.

  33. 33
    Hank Roberts says:

    >4, Spencer
    I see David Archer listed as a presenter; I hope you all will have more to say about this, here or at Stoat’s ‘Antarctic’ thread or somewhere else.

    Does anyone know where the current/daily ‘ice quake’ data shows up? The standard earthquake reports pick up smaller shakes but the online maps show Greenland as silent, which puzzles me.

  34. 34
    Hank Roberts says:

    Hmm, I didn’t know this:
    http://geohazards.cr.usgs.gov/staffweb/mcnamara/Pubs/papers_pdf/Columbia_IRIS.pdf

    “Alaskan and immediately adjacent Canadian glaciers supply one of the largest measured glaciological contributions to global sea level rise (~0.14 mm yr-1, equivalent to new estimates from Greenland). Retreating tidewater glaciers dominate the Alaskan sea level
    contribution due to their ability to efficiently transfer mass via iceberg calving [Arendt et al., 2002]. During retreat phase, a tidewater glacier may retreat on the order of 1-2 km yr-1 concurrent with dramatic increases in ice velocity…”

  35. 35
    Lynn Vincentnathan says:

    RE #5 “And now some number crunchers with dubious models and no real understanding yet as to how the climate really works are telling us what the climate will be like in a 100 years? There are so many events that could happen between now and then to change everything.”

    And one event between now and 100 years from now is that people might start reducing their GHGs, unlikely as that may seem. So, what d’ya say, let’s all reduce our GHGs by 70+% and end GW. That’ll show them arrogant, money-grubbing science types whose right.

    And BTW, who needs models to tell us that heat melts ice, though my Sunfrost frig seems to take forever to defrost. But once that ice gets to a certain point, big chunks just start dropping down, one after another, kaboom! kaboom!

    This brings me to catastrophe theory, which might fit those rather sudden changes in ice. I don’t know anything about CT, except that some functions look like potato chips….discontinuous, abrupt shifts.

  36. 36
    Walter Starck says:

    Since the first IPCC report in 1990 confidence expressed in the validity of the AGW hypothesis has increased with each successive report. The 1990 report stated that: “the observed (20th century temperature) increase could be largely due to… natural variability”. The 1995 report said: “the balance of the evidence suggests a discernible human influence on climate”. In 2001 it was claimed “there is new and stronger evidence that most of the warming observed over the last 50 years is attributable to human activities” and the current report concludes says it is: “90% probable” that the recent warming is “due to the observed increase in anthropogenic greenhouse gas concentrations”. During this time global expenditure on climate change research has been estimated to exceed $50 billion. It is now claimed that the science is settled,the evidence for AGW is overwhelming and the debate is over. . Unfortunately, few people, including most scientists other than climatologists, seem to have any clear or consistent idea of what this powerful new evidence actually is.

    It would be of great value to public understanding if a short list of perhaps five to ten of the most important new findings (since 1990) in support of AGW was available. I am not suggesting a detailed report, just an annotated list with a few links to key references. In view of overwhelming scientific evidence and consensus such a list should be a trivial task for the experts on this forum. Certainly it would provide a valuable clarification to the fog of claims and counterclaims in the public perception.

  37. 37

    It seems to me that when one has strong reason to suspect that one does not understand the most important factor in a situation (ice dynamics here), then quoting a quantitative value and a quantitative error range for all the second order effects which one does understand is singularly pointless. The reality is that climatologists currently have no adequate basis for quantitative projections of 21st century sea level, and they should say so as the main conclusion of the sea level section, not in hard-to-follow footnotes. The real story for the change from the TAR to AR4 is that “ice dynamics turned out to be much more important and rapid than we realized, and now we don’t know what to tell you” That should have been said in some reasonably clear manner.

  38. 38
    Ian K says:

    Stefan:
    You say you couldn’t get your suggested changes made to the summary. Why was this? Was it due to “political” pressures or time constraints, were you out-voted in some way, outgunned by more senior scientists, etc? Any further explanation of how the IPCC comes to a final document would be greatly appreciated by many of us outsiders!

  39. 39
    trevor says:

    Re #32: Have you checked the average temperatures at the South Pole and on top of the Greenland Ice Sheet? I think that you will find that even a 5deg C increase in annual average temperature will still see average temperature way below zero at the South Pole, and probably the same in Greenland.

    So tell me. If the average temperature is still below zero, how much ice will melt?

    [Response: Ice sheets are in equilibrium in a stable climate exactly because ablation (melt, in simple terms) balances accumulation (i.e. snowfall). In the center, where it is very cold, they gain mass, while around the edges they lose an equal amount. The problem with warming is not that it starts to melt the ice on the summit of Greenland or at the south pole, but around the edges of the ice sheets.]

    Second. Surely the issue relating to the mass of the ice sheet in the case of on-land ice sheets is precipitation as well as temperature. If temperatures are below zero, any precipitation will increase the ice volume. So precipitation must be a key factor. How much do we know about precipitation/temperature inter-relationships?

    [Response: As stated in my article, precipitation changes used in the projections are taken from a high-resolution atmospheric model. This is what leads to the assumption that Antarctica will gain mass overall, due to increased precipitation. However, this has not happened so far - until now, Antarctica seems to be losing mass.]

    Third. To assess the issue of rising sea levels, why not go to the coast when a spring high tide is acting. That will show you at least some of the effect.

    [Response: The problem of sea level rise does not arise on a calm day, even at spring tide. It arises during severe storm surges, which become a lot more frequent. For example, a study for New York showed that what is a once-in-a-century flooding event (submerging subway stations etc.) now, whould occur every 3 years if sea level were just 1 meter higher. -stefan]

  40. 40
    Alexander Ac says:

    James Hansen is even more “alarmistic” – the sea level rise of more than a 1 meter is quite well possible, read here:

    http://arxiv.org/ftp/physics/papers/0703/0703220.pdf

    Though I don’t know, if he attributes higher probability to +meter rise till 2100 than Stefan under BAU emmision scenario…

  41. 41
  42. 42
    Alan K says:

    First, thanks for the article – sea level rise is quite a hot topic right now. Secondly this post was not generated by a scep-bot.

    It’s understandable that you should focus on the upside of potential sea-level rise as that is what your models predict. However, to those of us who continue to wonder at the appropriateness of using models to predict climate, two of your comments stand out:

    “We therefore see that sea level appears to be rising about 50% faster than models suggest..”

    “A second problem with the above range is that the models used to derive this projection significantly underestimate past sea level rise.”

    Models surely are of less value if they are not used to produce the most accurate prediction of a future event. If they are wrong or have not reflected past events then the relevant parameters should surely be adjusted so that future projections will be consistent with past reality. Are you saying that they are and weren’t? “Significant” and “50%” are pretty fancy amounts wrong.

    If a model runs with a known bias, what is the value of that model? Further, if that bias only exists for you while the model builder (believes she) has it right, then that makes your model wrong. If the amount wrong a model can be (whether yours or hers) is 50% then you don’t need too many wrong models before you have one 100% wrong.

    I will anticipate the response that model results lie within a range (the IPCC stated one of which you in fact dismiss as being useful). But you have stated that they are wrong by so large a factor that one has to query how they are used, what outputs are released and of course more crucially what recommendations in terms of public policy are made as a result.

  43. 43
    Bill Tarver says:

    The whole point behind a summary is to – well, summarize. I accept that the IPCC has to be conservative and it is as well not to include very recent findings. However, we now seem to have a situation where the published worst case is actually not as bad as the true worst case. Politicians and journalists aren’t going to read the full report: the IPCC should update the summary when the report is published or risk an unwarranted degree of complacancy.

  44. 44
    Edward Greisch says:

    What is the maximum sea level rise before the ocean boils? I assume that there would be no ice anywhere at that time. The problem I have is that I don’t know the average slope of the shore. The shore is not vertical. What average shore slope is assumed in the IPCC computations? Please at least give me enough basic numbers to calculate backwards to get your assumed sea shore slope. Please send your answer to my email address.
    I conclude that I should not buy land within a few hundred feet of sea level. My very bad back of the envelope calculation says sea level cannot rise more than 800 feet before the ocean boils. If the ocean boils, we are all cooked anyway.

    [Response: If you melt all ice on Earth, sea level would rise about 70 metres. Last time this happened was in the Eocene, about 40 million years ago, when climate was about 4 ºC warmer than now (see Fig. 3.1-1 here).
    The IPCC does not assume an average shore slope, it gives vertical sea level rise. If you want to roughly see the effect on coastlines, you can do that interactively here. -stefan]

  45. 45
    pete best says:

    Due to the dynamics of ice being little understood why is it assumed by the IPCC (for one) that it is unlikely for Antartica or Greenland to melt significantly. Is it simply intertia or is there some other science such as thermodynamics telling us that they will not lose much ice between now and 2100?

    In additional how much ice would these places need to lose in order to weaken the oceans thermohaline/conveyor systems around the world ?

    The rate of ice decrease at these places is likely to stay relatively constant and hence not add to much to sea level rises because they are huge stores of ice that take millenia to weaken and wear down even with a large scale warming event taking place ?

  46. 46
    Fergus Brown says:

    For the past month or so, sea surface temperature anomalies in the areas between the Antarctic and New Zealand and, subsequently, the Pacific, have been strongly negative. This has happened before, recently. ‘Tracking’ the path of the anomalies indicates their source.
    In addition, I note this new paper in GRL: ‘Rapid Freshening of Antarctic Bottom Water from the Indian and Pacific Oceans’; Rintoul, March 2007.
    It seems clear that the SSTAs are a function of seasonal sea ice loss at the surface. Does the Rintoul paper suggest the possibility of large outflowing of freshwater from below the surface, perhaps from subglacial lakes or rivers?
    Can you tell me if either of these phenomena are unusual and, if so, whether such processes are accurately reflected in current estimates of ice mass balance, or whether they, too, suggest an underestimate. The Rintoul paper, in particular, may be indicative of a process which has not, afaik, previously been considered.
    Regards,

  47. 47
    Dick Veldkamp says:

    Re #40, #41 Hansen’s paper

    Stefan, would you care to comment on Hansen’s paper? It seems to me that he makes a good case for “more alarmism”, i.e. scientist must speak out and say that we should seriously consider a sea level rise >1 m on a century scale, because there is a non-negligible possibility of that happening. From your article I guess you would agree that there is a 1% possibility (say). Wouldn’t that justify raising the alarm?

    I think there is another reason why we should. The reaction I see in newspaper articles etc is: ‘Oh, IPCC says it’s just going to be 20-60 cm by 2100 [and Al Gore is an alarmist for bringing up the possibility of >5 m rise]. Like there will be some magical cut off in 2100! Of course history does not end in 2100, so if we’re not going to see large sea level rise in the 21st century, that’s no comfort at all for our grandchildren.

    I am aware of the “crying wolf problem”; this is a complicated issue, and I don’t want to advocate “all out alarmism” just now, but I would like to hear some opinions.

    [Response: I did not know Hansen's paper but read it just now. I fully agree with what he writes about "scientific reticence", his words echo my own experience very well. In many IPCC discussions I have noticed a strange asymmetry: people were very concerned about possibly erring on the high side (e.g., the upper bound of sea level rise possibly being criticised as "alarmist"), and not very concerned about erring on the low side (or some even regarding this as a virtue of being "cautious"). How likely is it actually that the rate of sea level rise in this century would on average be only half of the rate currently observed, despite further warming? But this is what would have to happen to reach the lower end of the IPCC range, namely 18 cm. (Current observed rate is 3.1 mm/year according to IPCC, or 3.3 mm/year using the satellite data 1993-2006.) Nobody was very concerned that 18 cm is a rather implausibly low value, possibly related to the fact that the models used to produce it already greatly underestimate the past sea level rise. Imagine the reverse had happened: models that greatly overestimate past sea level rise and come up with some implausibly high sea level rise number. Would IPCC have simply published that, as they did with the 18 cm value? I very much doubt it. Giving a low value is considered "safe", it requires no courage for sticking your neck out, while giving a high number is considered risky and alarmist. I don't think we are doing our job properly if we apply double standards to "low" and "high" estimates in this way. We need to dispassionately look at all the evidence, regardless of what is politically convenient or risky. -stefan]

  48. 48
    Tom Huntington says:

    Stefan: Very informative post, thank you. I wanted to ask if you could give a citation for the study you mentioned in your response to post No. 39 about frequency of subway flooding in New York.

    [Response: Rosenzweig, C and Solecki, W D (eds) (2001) Climate Change
    and a Global City: The Potential Consequences of Climate
    Variability and Change. Metro East Coast. Report for the US
    Global Change Research Program. National Assessment of the
    Potential Consequences of Climate Variability and Change for
    the United States. Columbia Earth Institute, New York. ]

    I find your figure comparing the IPCC 2001 SLR scenarios with observed SLR to be compelling. The question I hope the appropriate scientific community is addressing is why is sea level rising 50% faster than the modeling projected? What are the most likely reasons for this? Reading your post and following some of the literature on dynamical changes it is tempting to conclude that the difference may be due to acceleratded ice sheet flow in Greenland and Antarctica that is not accounted for in the models. But there could also be errors in the modeling of the melt rate of mountain glaciers, the rate of snowfall on Greenland and Antarctica, and the underdog, failure to account for significant net groundwater depletion.

    It might be interesting to show observed SLR in comparison to projections of the 1st and 2nd IPCC projections as well.

  49. 49
    Dan Fregeau says:

    Per the IPCC comment noted above:

    “Contraction of the Greenland ice sheet is projected to continue to contribute to sea level rise after 2100. … The corresponding future temperatures in Greenland are comparable to those inferred for the last interglacial period 125,000 years ago, when paleoclimatic information suggests reductions of polar land ice extent and 4 to 6 m of sea level rise. {6.4, 10.7}”

    Further interglacial sea level rise is probably inevitable irrespective of anthropogenic warming contributions. Why should warming during the current Holocene Interglacial be significantly different/less than during the previous Eemian Interglacial (MIS5e), when sea levels were, per the IPCC’s own comments, above modern levels or for that matter above the purported mid-Holocene highstand?

    In fact, the literature is pretty clear that the better analog for the Holocene is actually MIS11 (Droxler), around 400 kya, when orbital eccentricity was quite low as it is today, resulting in a longer interglacial (Berger) with sea levels higher than during the Eemian.

    Note that during both the Eemian and MIS11, CO2 levels were lower than they are today, which strongly suggests a polar warming/melting mechanism other than CO2. Perhaps this anthropogenic warming debate is an unfortunate distraction from what we should be really focusing on – preparing for higher interglacial warming and rising sea levels, period.

  50. 50
    George K says:

    Let’s put some facts on the table:

    - The poles go through (nearly) 6 months of darkness each year. With no sunlight in the winter, and with the sun at a low angle in the sky in the summer, it is always going to be cold at the poles and any melting in the summer is going to freeze back in the winter;

    - The average annual temperature at the south pole is -49.5C. The average annual temperature at the north pole is -25.0C. 5.0C of warming leaves the poles very cold.

    - Sea level has been rising at 1mm to 3mm every since the large continental glaciers melted after the last ice age – ie for the last 9,000 years.

    - The interior of Greenland is below sea level. This is important if one is thinking of all of the ice melting or one is thinking of the glaciers “sliding off” into the sea.

    - The arctic is still experiencing “rebound” from the weight of the glaciers from the last ice age.

    - The Minimum Sea Ice Extent in the arctic was lower in 1990 than in 2006 – ie the arctic ice summer (September) minimum has been more-or-less stable for 16 years.

    [Response: I'm always suspicious when someone announces: now here come the facts! You'd have to give some references for some of your "facts". Sea level has been rising for 9,000 years at the rate of 1-3 mm/yr? So in the Middle Ages it was 1-3 meters lower than now? A number of studies (interestingly, some looking at where the Romans built fish ponds and other structures connected to the sea) and the IPCC rule this out, concluding just the opposite: in the preceding millennia, there was not even remotely the rate of sea level rise that is observed for the last century, it is a modern phenomenon. And on the September arctic sea ice cover, NASA has a very different story. -stefan]


Switch to our mobile site